Skip to main content

Full text of "Practical marine engineering for marine engineers and students, with aids for applicants for marine engineers' licenses"

See other formats


Practical Marine Engineering 




Aids for Applicants for Marine Engineers' Licenses 



New York 

Marine Engineering, Inc. 

309 Broadway 




Copyright, 1901 


A I \HE purpose of the author in the preparation of this work 
has been to provide help for the operative or practical 
marine engineer, either for the man who has already en- 
tered the profession, but who may wish to perfect himself more 
fully in many branches of the subject, or for the applicant for the 
lowest round of the ladder, or for the young man whose atten- 
tion is first turning to this field, and who may wish some simple 
and fairly complete presentation of the subject from the prac- 
tical standpoint. 

The treatment of the subject throughout has thus been with 
a view to simplicity, but without undue sacrifice of generality or 
exactness of statement. It has been the desire of the author to 
bring the subject, so far as treated in the present work, within 
the grasp of those who have not had the advantages of higher 
mathematical and engineering education, but who may wish, 
nevertheless, to fit themselves for positions of honor and respon- 
sibility in the field of operative marine engineering. 

With this end in view only such parts of the general field of 
engineering have been included as are of special interest to 
the practical marine engineer. On these topics, however, the 
attempt has been made to give the largest amount of useful in- 
formation in the simplest and most compact form. In the ma- 
rine field itself likewise, selection has been necessary, and 
many interesting parts of the subject have been omitted or 
briefly referred to in order to give more room for the practical 
side of the subject. Thus the book does not treat of the de- 
signing of marine machinery except in an incidental way. For 
the operative engineer the topics of greater importance are con- 
struction, operation, management and care. The simpler parts 
of the subject of design are, however, represented by the U. S. 




rules regarding the design and construction of marine boilers, 
and by many hints regarding proportions and relations scat- 
tered throughout the work. 

In the chapters dealing descriptively with engines, .boilers 
and auxiliaries, it has been impossible of course, to describe ex- 
haustively every form of design or appliance to be met with 
in marine practice. The purpose has been rather to describe 
typical or standard forms, and to give the general conditions 
which the various parts must fulfil. The illustrations have been 
specially chosen with a view to supplement the text in these 
various particulars, and it is hoped that they will form not the 
least instructive and acceptable feature of the work. 

The subject of operation, management and repair has been 
given special attention, and it is hoped that this part of the work 
will be of value, especially to the young engineer lacking in 
practical experience. 

In Chapter VIII is gathered a collection of miscellaneous 
problems and discussions, many of which, it is hoped, will be of 
value to the professional engineer in connection with the vari- 
ous questions likely to arise in his experience. The chapters 
on valve gears and on indicator cards while necessarily brief are 
intended to present the fundamental features of the subject in 
such manner as to aid the novice and instruct and stimulate 
the professional engineer to a better understanding of these im- 
portant branches of the subject. The chapter on propulsion 
and powering is necessarily brief, but the fundamental principles 
are given, with a few simple rules and the discussion of most of 
the problems commonly arising in practical engineering work. 

The chapters on refrigeration and on electricity on shipboard 
are added in order to give the marine engineer some notion of 
the fundamental principles controlling the operation of refrig- 
erating and electric machinery, these two important auxiliaries 
of modern marine engineering practice. They are of necessity 
quite incomplete, especially Chapter XII, but it is hoped that 
nevertheless they may be of aid to the marine engineer in un- 
derstanding the mode of operation of such machinery, and in 
giving to it the proper care. 

In Part II is given an elementary discussion of computa- 
tions for engineers, or rather of the mathematics upon \vhich 
such computations depend. A general knowledge of the sub- 
ject is pre-supposed, but the more essential features of the ele- 



mentary mathematics usually required are given and illustrated 
with many problems. It is hoped that this feature of the work 
may be of aid to those who wish to drill themselves in such com- 
putations as the marine engineer is commonly called upon to 

Attention may also be called to the appendix containing a list 
of questions, each with page reference to the part of the book 
where the answer may be found. The answer, of course, will 
not usually be found in the direct form suggested by the ques- 
tion, but a discussion of the subject will be found giving the in- 
formation needed for the answer, which may be put into form 
by the reader for himself. It is believed that such an exercise 
will be of far greater value than the perusal of a series of ques- 
tions and answers in the usual catechism form. 

Throughout the work numerous problems have been scat- 
tered, accompanied usually by illustrative examples, showing 
the method of working. Numerous cross references have also 
been given, to aid in the more complete explanation of any 
given topic, and in finding these use should be made of the table 
of contents, giving the page location of each section and bracket 

A collection of miscellaneous problems is also added at the 
end of the book, as well as a set of steam tables for use in the so- 
lution of the various problems requiring a knowledge of its vari- 
ous mechanical and physical properties. 




Principal Materials of Engineering Construction. 


1. Aluminum I 

2. Antimony I 

3. Bismuth 2 

4. Copper 2 

5. Iron and Steel 3 

i ] Cast Iron 4 

2] Malleable Iron 9 

3 Wrought Iron 10 

4 Blisters and Laminations u 

5 Steel 12 

6. Lead 23 

7. Tin 24 

8. Zinc 24 

9. Alloys 24 

10. The Testing of Metals ' 26 

i ] Different Kinds of Tests 26 

2] Explanation of Terms Used 26 

3] Test Pieces for Iron 28 

4 Test Pieces for Steel and Other Materials 28 

5] Bending, Quenching and Hammer Tests 29 



11. Coa! 3i 

i] Composition and General Properties 3 1 

2] Combustion 3 2 

3] Impurities in Coal. Clinker Formation 35 

4] Weathering of Coal 36 

5] Spontaneous Combustion 37 

6] Corrosion 4 

7! Transportation and Stowage \ 40 

8] General Comparison Between Bituminous and Anthracite 

Coal 4i 

12. Briquettes and Artificial Fuel 41 

13. Liquid Fuel 4 2 

i ] Composition 42 

2] Combustion 43 

"3! Danger of Explosion : 44 

4] Evaporative Power 44 

"5l Stowage and Handling 45 

6] Use of Oil and Coal Combined 46 

7] Cost 46 






14. Types of Boilers 47 

i ] The Scotch Boiler 50 

2] Direct Tubular Boiler, Gunboat Type 50 

Direct Tubular Boiler, Locomotive Type 51 

The Flue and Return Tubular or Leg Boiler 51 

5J The Flue Boiler 52 

: 6 n Water-Tube Boilers 53 

7] Relative Advantages of Different Types of Boilers 56 

15. Riveted Joints 67 

16. Materials and Construction 86 

i ] Materials 86 

2] Joints 86 

3] Construction of Fire-Tube Boilers 87 

4] Construction of Water-Tube Boilers . 108 

5] Common Sizes and Dimensions of Scotch Boilers 112 

'6] Common Proportions for Scotch Boilers 112 

7] Weights of Boilers 112 

8 Western River Boat or Flue Boilers 113 

17. Boiler Mountings and Fire Room Fittings 115 

i] Safety Valves IIS 

2] Muffler 118 

3] Stop Valve n8 

4] Dry-Pipe, or Internal Steam Pipe 120 

'5 Feed Check Valve and Internal Feed Pipe 121 

Surface and Bottom Blows 122 

7 Steam Gauges 124 

Water Gauge and Cocks 125 

[9] Hydrokineter 127 

[10] Hydrometer 127 

[III Boiler Saddles 128 

[12 Boiler Lagging 129 

18. Draft 129 

19. Boiler Design in Accordance with the Rules of the U. S. Board 

of Supervising Inspectors of Steam Vessels , 138 


Marine Engines. 

20. Types of Engines and Arrangement of Parts 154 

21. Description of Principal Parts of a Marine Engine 165 

[i] Cylinders 165 

"2] Columns 168 

3 1 Bed-Plates 172 

'4! Engine Seating 174 

5] Pistons 175 

6] Piston-Rods 179 

7] Crossheads 180 

8] Connecting-Rods 184 

9] Crank Shafts 185 

[10] Line Thrust and Propeller Shafts 188 

[nl Bearings 190 

22. Western River Practice 199 

23. The Steam Turbine 207 

24. Engine Fittings 209 

f i ] Throttle Valve 209 

\2\ Main Stop Valve 212 

[3] Cylinder Drain Gear and Relief Valves 215 



4 Starting Valves 216 

5] Reversing Gear 217 

6] Turning Gear 220 

7] Joints and Packing 220 

Reheaters 225 

9J Governors 225 

[ 10] Counter Gear 227 

fii] Lagging 228 

[12] Lubrication and Oiling Gear 228 

25. Piping 237 

[i] Systems and Materials 237 

\2\ Expansion Joint 239 

[3] Globe Angle and Straightway Valves 240 



26. Circulating Pumps '. 241 

27. Condensers 243 

28. Air Pumps 245 

29. Feed Pumps and Injectors 251 

30. Feed Heaters 255 

31. Filters 260 

32. Evaporators 261 

33. Direct Acting Pumps 263 

34. Blowers or Fans 267 

35. Separators 267 

36. Ash Ejectors 269 

37. General Arrangement of Machinery 271 


Operation, Management and Repair. 

38. Boiler* Room Routine 273 

[i] Starting Fires and Getting Under Way 273 

[2] Fire Room Routine 277 

[3] Emergencies and Casualties 287 

39. Engine-Room Routine and Management 298 

[i] Getting Under Way 298 

[2] Routine Operation 301 

[3] Minor Emergencies and Troubles 304 

40. Boiler Corrosion 308 

41. Boiler Scale 319 

42. Boiler Overhauling and Repairs 328 

i ] Inspection and Test 328 

2] Leakage from the Joints of Boiler Mountings 333 

'3] Leakage About Shell Joints 333 

4] Leakage at Internal Joints 334 

5] Patches 336 

61 Cracks and Holes 336 

7 Blisters and Laminations 337 

8] Tubes 338 

[9! Leakage About Stays and Braces 339 

[10] Bulging or Partial Collapse of Furnace or Combustion 

Chamber Plates 339 

FIT] Split in Feed-Pipe 341 

43. Engine Overhauling, Adjustment and Repairs 341 

i] Cylinders 342 

2] Pin Joints and Bearings 343 

3] Crosshead Guides 345 

4] Crosshead Marks 346 




Lining Up 347 

Valve Gear 352 

Thrust Bearings 353 

Circulating Pump 353 

Condensers 353 

Air Pumps 355 

Pumps in General 355 

Piping 355 

44. Spare Parts 356 

45. Laying Up Marine Machinery 356 


Valves and Valve Gears. 

46. Slide Valves 358 

i] Simple Slide Valve 358 

2] Double Ported Slide Valve 360 

3J Piston Valve 361 

4] Equilibrium Piston 365 

5] Jy' s Assistant Cylinder 366 

6] Equilibrium Rings 367 

_7] Outside and Inside Valves 368 

47. Motion Due to Simple Excentric and Its Representation by 

Valve Diagram 369 


Simple Excentric 369 

Oval Valve Diagram 372 

Bilgram Valve Diagram 377 

Zeuner Valve Diagram 379 

48. Stephenson Link Valve Gear 380 

49. Braemme-Marshall Gear 386 

50. Joy Valve Gear 391 

51. Walschaert Valve Gear 391 

52. Crank Valve Gear 393 

53. Details of Stephenson Link Valve Gear * 397 

fi] Excentric and Strap and Excentric Rod 397 

[2] Link 400 

[3] Link Block and Valve Stem 402 

54. Valve Setting 405 

i] Putting an Engine on the Center 405 

[2] Setting the Valve 406 

[3] Valve Setting from the Indicator Card 408 



Steam Bngine Indicators and Indicator Cards. 

55. Indicator Cards 410 

[i] Descriptive 410 

[2] The Indicator Card and the Operation of the Valve Gear. . 413 

[3] Working up Indicator Cards for Power . 417 

[4] Combined Indicator Cards 425 

56. Steam Engine Indicators 429 

[i] Descriptive 429 

[2] Reducing Motions 431 

[3] Taking an Indicator Card 434 


Special Topics and Problems. 

57. Heat and the Formation of Steam 438 

[i] Constitution of Matter 438 

[2] Heat 439 




[3] Steam ; 444 

[4] Total Heat in a Substance 450 

[5] Latent Heat in Passing from Ice to Water 452 

58. Steam Boiler Economy 453 

T ] General Principles 453 

[2] Evaporation Per Pound of Coal 455 

13] Evaporation Per Pound of Combustible 459 

59. Steam Engine Economy 460 

[i] General Principles 460 

[2] Relation of Expansion to Economy 469 

[3] Economy of the Actual Engine 471 

60. Coal Consumption and Related Problems 472 

61. The Lever Safety Valve and the Safety Valve Problem 476 

62. The Boiler Brace Problem 480 

63. Strength of Boilers 485 

64. Loss by Blow Off 489 

65. Gain by Feed Water Heating 491 

66. The Proportions of Cylinders for Multiple Expansion Engines. . 492 

67. Clearance and Its Determination 494 

68. The Iiffect of Clearance in Modifying the Apparent Expansion 

Ratio as Given by the Point of Cut-Off 496 

69. Engine Constant 497 

70. Indicated Thrust 498 

71. Reduced Mean Effective Pressure 499 

72. Pressure on Main Guides 502 

73. Force Required to Move a Slide Valve 503 

74. Amount of Condensing Water Required 504 

75. Work Done by Pumos 505 

76. Discharge of Steam Through an Orifice 507 

77. Computing Weights of Parts of Machinery 508 

FT] Units to be Used ". 508 

[2] Approximations and Short Cuts 509 


Propulsion and Poweringf. 

78. Measure of Speed 514 

79. Propulsion 514 

80. Screw Propeller 517 

[i] Definitions 517 

[2] Varieties of Propellers 523 

[3! Materials 526 

f4l Measurement of Pitch 526 

81. Paddle Wheels 531 

82. Powering Ships 534 

83. Reduction of Power When Towing or When Vessel is Fast 

to a Dock 537 

84. Trial Trios 539 

85. Special Conditions for Speed Trials 544 



86. General Principles 545 

87. Refrigeration by Freezing Mixtures 546 

88. Refrigeration by Vaporization and Expansion 547 

89. Principal Features of Ammonia Refrigerating Apparatus 549 

90. Refrigeration by the Expansion of a Compressed Gas 553 

91. Principal Features of Compressed Air Refrigerating Apparatus 554 

92. Operation and Care of Refrigerating Machinery 556 




Electricity on Shipboard. 


93. Introductory 559 

94. The Dynamo 565 

95. Wiring and The Distribution of Light and Power 571 

96. Lamps 575 

97. Operation and Care of Electrical Machinery 577 

i ] Routine Care 577 

'2] Faults 578 

Part II. 



1. Common Fractions 581 

i] Units of Measurement and Definitions 581 

2] Reduction of a Mixed Number to an Improper Fraction.. 583 

3] Reduction of an Improper Fraction to a Mixed Number 584 

4] Reduction of Fractions Without Change of Value 584 

5] Addition of Common Fractions 586 

6] Subtraction of Fractions 587 

7] Multiplication of Fractions 588 

8] Divisions of Fractions 588 

9] Multiplication and Division of Fractions 589 

[10] Complex Fractions 591 

2. Decimal Fractions 592 

i ] Introductory 592 

2] To Reduce Decimals to Lower Terms 593 

3 To Raise Decimals to Higher Terms 593 

4] To Reduce a Decimal Fraction to a Common Fraction. .. . 593 
51 To Reduce a Common Fraction, Proper or Improper, to a 

Decimal 593 

6] To Add Decimals 594 

7] To Subtract Decimals 594 

8] To Multiply Together Two Numbers Expressed Decimally 595 
9] To Find the .Quotient of Two Quantities, Expressed Deci- 
mally 595 

3. Percentage - 596 

4. Compound Numbers 599 

[ 1 1 Long or Linear Measure 599 

'2] AvoirdtiDois Weight or Measure 599 

"3] Square Measure 599 

4] Cubic or Volume Measure 600 

"5] Liquid Measure 600 

6] Dry Measure 600 

7l Shipping Measure 600 

8] The Metric System of Weights and Measures 600 

[9! Conversion Tables 601 

[iol Reduction of Compound Numbers 602 

fill Addition of Compound Numbers 603 

[12] Subtraction of Compound Numbers 603 

[13] Multiplication of Compound Numbers 604 

[14] Division of Compound Numbers 604 

5- Duodecimals 605 

6. Ratio and Proportion 607 

[i] Simple Proportion 607 



[2] Compound Proportion 610 

7. K\ oiution and Involution 612 

f i ] Introductory 612 

[2] ' To Extract the Square Root 613 

[3] To Extract the Cube Root 615 

8. Mathematical Signs, Symbols and Operations 616 

9. Geometry and Mensuration 621 

i ] Square 621 

2] Rectangle 622 

3] Parallelogram 622 

4] Trapezoid 623 

5] Triangle 623 

6 Right-Angled Triangle 624 

7 j Trapezium 625 

81 Regular Polygons 625 

9] Irregular Figures 626 

10] Circle 626 

1 1 Circular Ring or Annulus 628 

12] Sector of Circle 628 

13] Segment of Circle 629 

14] Ellipse 629 

15] Figures With an Irregular Contour 630 

16] Prism 633 

17] Cylinder 634 

[18] Any Solid with a Constant Section Parallel to the Base, 

Either Right or Oblique 635 

[19] Wedge 635 

[20] Right Pyramid * 635 

\2i General Pyramid 636 

[22 Right Circular Cone 637 

[23 General Cone 637 

[24] Frustum of Right Pyramid '. . 638 

[25] Frustum of General Pyramid 639 

T26] Frustum of Right Cone 639 

[27] Frustum of General Cone 640 

[28] Sphere 640 

[29! Volume of Irregular Shape 640 

[30 Volume Generated by Any Area Revolving About an Axis 641 

10. Problems in Geometry 642 

[i] At Any Point in a Straight Line to Erect a Perpendicular 642 

[2] To Bisect the Distance Between Two Points 643 

[3] To Find the Center from which to Pass an Arc of Given 

Radius Through Two Given Points 643 

[4] To Divide a Given Line into a Given Number of Equal 

Parts 643 

[5] To Construct a Triangle, Having Given the Three Sides. .. 643 

[6] To Bisect a Given Arc or Angle 644 

[7] To Construct a Mean Proportional 644 

"[8] To Construct a Fourth Proportional 644 

[9] To Construct a Square Equivalent in Area to a Given 

Rectangle 645 

[10] To Construct a Square Equivalent in Area to a Given 

Triangle 645 

TII] With One Given Side, to Construct a Rectangle Equivalent 

to a Square 64; 

[12! To find the Length of an Arc of a Curve 645 

[13] To Construct an Ellipse 646 

[14] To Construct any Regular Polygon 647 

[15] To Develop the Surface of a Cylinder 648 



[16] To Develop the Surface of a Cylinder which is Intersected 
by Another Cylinder, the Two Axes being in the same 

Plane 648 

~i?] To Develop the Surface of a Cone t . . . . 649 

18] To Develop the Surface of the Frustum of a Cone 649 

[19] To Develop the Segments of an Elbow 649 

11. Physics , 650 

i] Acceleration Due to Gravity 650 

2] Specific Gravity 650 

3], Heat Unit 651 

4] Specific Heat 651 

5] Expansion of Metals 651 

12. Mechanics 652 

1 Introductory 652 

2 Force 652 

3 Specification of a Force 652 

4 Moment of a Force 653 

5] Resultant 653 

6] Work 653 

7] Power 653 

8] Energy 655 

[9] Conservation of Energy 656 

loj Statics 656 

1 1 ] Dynamics 656 

12] Propositions in Statics 656 

13] Mechanical Powers 659 

14] Examples in Mechanics 665 


Practical Marine Engineering 



Sec. i. ALUMINUM. 

The commercially pure metal, i. e., with less than i per cent 
impurity, is white in color, soft, ductile and malleable. It melts 
at about 1,160 F., has a tensile strength of about 15,000 Ib. per 
square inch of section, but lacks in stiffness and resilience, or the 
power to withstand shocks. 

Aluminum does not oxidize readily under the influence of 
ordinary air, but when in contact with sea water, or in air 
charged with sea water, the corrosion is often serious in extent. 
Aluminum cannot be welded except electrically, is not suitable 
for forging or rolling when hot, and cannot be tempered or hard- 
ened. It is, however, suitable for casting, and when cold can be 
rolled into sheets and drawn into wire, and in thin sheets or small 
pieces may be spun or flanged, or worked under the hammer in 
various ways. 

Aluminum unalloyed is of comparatively small value to the 
engineer, but it enters into several valuable alloys, as described 
in Section 9, and its use in this way has increased to a consider- 
able extent within the past few years. 

Sec. 2. ANTIMONY. 

The pure metal is whitish in color, quite brittle and crystal- 
line or laminated in structure, and has a melting point of about 
840 F. It is useless in the pure state for ordinary engineering 
purposes, but is a valuable ingredient of various alloys used for 
bearing metals, etc., as described in Section 9. 


Sec. 3. BISMUTH. 

The pure metal is light red in color, very brittle and highly 
crystalline in structure, with a melting point of about 510 F. It 
is useless in the pure state for engineering purposes, but forms a 
part of various alloys used for bearing metals, etc., as described 
in Section 9. 

Sec. 4. COPPER. 

In the pure state the metal is red in color, soft, ductile and 
malleable, with a melting point of about 2,000 F., and a tensile 
strength of from 20,000 to 30,000 Ib. per square inch of section. 
Copper is not readily welded except electrically, but, on the 
other hand, is readily joined by the operation of brazing. At- 
tempts have been made to temper or harden it, but the operation 
has not been made a practical success. It is readily forged and 
cast, and when cold may be rolled into sheets or drawn into wire, 
and in sheets or small pieces may be spun or flanged or worked 
under the hammer in various ways. 

The tensile strength of copper, rapidly falls off as the tem- 
perature rises above about 400 F., so that at from 800 to 900 the 
strength is only about one-half what it is at ordinary tempera- 
tures. This peculiarity of copper should be borne in mind when 
it is used in places where the temperature is liable to rise to these 
figures. Again, if copper is raised nearly to its melting point in 
contact with the air it readily unites with oxygen and loses its 
strength in large degree, becoming, when cool, crumbly and brit- 
tle. Copper in this condition is said to have been burned. The 
possibility of thus injuring the tenacity of copper is of the highest 
importance in connection with the use of brazed joints in steam 

In the operation of brazing a joint, the surfaces to be joined 
are cleaned, bound together with wire or otherwise, then sup- 
plied with brazing solder in small bits, mixed with borax as a 
flux, and placed in a clear fire until the solder melts and forms 
the joint. The brazing solder, or hard solder, as it is often called, 
is usually a brass or alloy of copper and zinc. The melting point 
of all such alloys is below that of copper, and when copper is 
joined to brass, or two pieces of brass are joined together, the 
solder used must have a melting point lower than either of these 
metals. In the operation of brazing a copper joint, therefore, the 
greatest care must be taken in the selection of a solder and in 
attention to the fire, so that there may be no danger of burning 


the copper, and thus endangering the quality of the metal in the 

Copper unalloyed is used chiefly for pipes and fittings, espe- 
cially for junctions, elbows, bends, etc. For large sizes the ma- 
terial is made in sheets, bent and formed to the desired shape and 
brazed at the seams. Small sizes are either made by the same 
general process or from solid drawn pipe, which may be bent as 
desired after drawing. Copper is also largely used as the chief 
ingredient of the various brasses and bronzes, as described in 
Section 9. 



It will be convenient to give here a general classification of 
iron and steel products based on the methods of manufacture. 
The following is the classification used by Prof. J. B. Johnson in 
his text book on the Materials of Construction. 


Wrought Iron. Rolled or forged from a puddle ball ; it con- 
tains slag and other impurities and cannot be hardened by sud- 
den cooling. 

Steel. Rolled or forged from a cast ingot and free from slag 
and similar matter. 

Soft Steel. Will weld (with care), and cannot be hardened 
by sudden cooling. It is sometimes called ingot iron, and has the 
same uses as wrought iron. 

Medium Steel. Welds imperfectly except by electricity. 
Will not harden by sudden cooling. 

Hard Steel. Will not weld. Hardens by sudden cooling. 
Tool steel, etc. 


Steel Castings. Malleable metal cast into forms. 
Malleable Cast Iron. Non-malleable metal cast into forms 
and then brought to a semi-malleable condition. 


Cast Iron ; Hard Cast Steel. Non-malleable metal cast into 

In describing these products at length we shall find it con- 
venient to begin with cast iron. 


[i] Cast Iron. 

This material consists of a mixture and combination of iron 
and carbon, with other substances in varying proportions. 

(i) Influence of Carbon. In the molten condition the car- 
bon is dissolved by the iron and held in solution just as ordin- 
ary salt is dissolved by water. The mixture or combination of 
the two elements is thus entirely uniform. The proportion of 
carbon which pure melted iron can thus dissolve and hold in 
solution is about 3^ per cent. If chromium or manganese is 
present also, the capacity for carbon is much increased, while 
with silicon, on the other hand, the capacity for carbon is de- 
creased. In the various grades of cast iron the proportion of 
carbon is usually found between 2 per cent and 4.5 per cent. 

Now, when such a molten mixture cools and becomes solid, 
there is a tendency for a part of the carbon to be separated out 
and no longer remain in intimate combination with the iron. 
The carbon thus separated or precipitated out from the iron 
takes that form known as graphite, and collects together in very 
small flakes or scales. The carbon which remains in intimate 
combination with the iron is said to be combined, while that 
which is separated out is usually called graphitic. 

The qualities of cast iron depend chiefly on the proportion 
of total carbon and on the relative proportions of combined and 
graphitic carbon. 

With a high proportion of graphitic carbon the iron is soft 
and tough, with low tensile strength, and breaks with a coarse 
grained dark or grayish colored fracture. In fact the substance 
in this condition may be considered as nearly pure iron with fine 
flakes of graphite entangled and distributed through it, thus giv- 
ing to the iron a spongy structure. The iron thus forms a kind 
of continuous mesh about the graphite, which decreases the 
strength by reason of the decrease of cross-sectional area 
actually occupied by the iron itself. Such irons are termed gray. 

As the relative proportion of graphitic carbon decreases 
and that of combined carbon increases, the iron takes on new 
properties, becoming harder and more brittle. Its tensile 
strength also increases to a certain extent, and the fracture be- 
comes fine grained or smooth and whiter in color. When these 
characteristics are pronounced the iron is said to be white. 
When about half the carbon is combined and half separates out 
as graphite, the effect is to produce a distribution of dark spots 


or points scattered over a whitish field. Such irons are said to 
be mottled. 

In a general way with a large proportion of total carbon 
there is likely to be formed a considerable amount of graphitic 
carbon, and hence such irons are usually gray and soft. With 
a large proportion of carbon also the iron melts more readily and 
its fluidity is more pronounced. As the proportion of total car- 
bon decreases the cast iron approaches gradually the condition 
of steel, whose properties will be discussed in a later paragraph. 

Of the special ingredients in cast iron the. combined carbon 
is one of greatest importance. It is that chiefly which by uniting 
with the iron gives it new qualities, and the principal influence 
of other substances lies in the effect which they may have on the 
proportion of this ingredient. As between graphitic and com- 
bined carbon, the former does not affect the quality of the iron 
itself, but acts physically by affecting the structure of the cast- 
ing; while the latter, by entering into combination with the 
iron, acts chemically and produces a new substance with different 
qualities. The following percentages of combined carbon are 
recommended for qualities of iron as indicated: 


Soft cast iron 10 to . 15 of one per cent. 

Greatest tensile strength about .45 of one per cent. 

Greatest transverse strength about . 70 of one per cent. 

Greatest crushing strength one per cent or over. 

The proportions of combined and graphitic carbon are in- 
fluenced by the rate of cooling, and by the presence or absence 
of various other ingredients. Slow cooling allows time for the 
separation of the carbon and thus tends to form graphitic car- 
bon and soft gray irons. Quick cooling, or chilling in the ex- 
treme case, prevents the formation of graphitic carbon and thus 
tends to form hard, white irons. 

In addition to carbon, small quantities of silicon, sulphur, 
phosphorus, manganese and chromium may be found in cast iron. 

(2) Influence of Silicon. The fundamental influences of 
silicon are two. (a) It tends to expel the carbon from the com- 
bined state and thus to decrease the relative proportion of com- 
bined carbon and increase that of graphitic carbon. (b) Of 
itself silicon tends to harden cast iron and to make it brittle. 

These two influences are opposite in character, since an in- 
crease in graphitic carbon softens the iron. In usual cases the 


net result is a softening of the iron, an increase in fluidity, and 
a general change toward those qualities possessed by iron with 
a high proportion of graphitic carbon. This applies with a pro- 
portion of silicon from 2 per cent to 4 per cent. With 
more than this the influence on the carbon is but slight and the 
result on the iron is to decrease the strength and toughness, 
giving a hard but brittle and weak grade of iron. 

A chilled cast iron is an iron which if cooled slowly would 
be gray and soft, but, as explained in (i), by sudden cooling, 
from contact with a metal mould or other means, becomes white 
and hard, especially at and near the surface. Certain grades 
of cast iron tend to chill when cast in sand moulds. This prop- 
erty is usually undesirable. In such cases the tendency can be 
prevented by the addition of silicon, which, by forcing the car- 
bon into the graphitic state on cooling, prevents the formation 
of the hard, chilled surface. In all cases the actual effect of 
adding silicon will depend much on the character of the iron 
used as a base, and only a statement of the general tendencies 
can here be given. 

To sum up, a white iron which would give hard, brittle and 
porous castings can be made solid, softer and tougher by the ad- 
dition of silicon to the extent of perhaps 2 or 3 per cent. As the 
silicon is increased the iron will become softer and grayer and 
the tensile strength will decrease. At the same time the shrink- 
age will decrease, at least for a time, though it may increase 
again with large excess of silicon. The softening and toughen- 
ing influence, however, will only continue so long as additional 
graphite is formed, and when most of the carbon is brought into 
this state the maximum effect, has been produced, and any fur- 
ther addition of silicon will decrease both strength and tough- 

(3) Influence of Sulphur. Authorities are not in entire 
agreement as to the influence of sulphur on cast iron, some be- 
lieving that it tends to increase the proportion of combined car- 
bon, while others maintain that it tends to decrease both the 
combined carbon and silicon. It is generally agreed, however, 
that in proportions greater than about .15 to .20 of I per cent it 
increases the shrinkage and the tendency to chill, and decreases 
the strength. Sulphur does not, however, readily enter cast 
iron under ordinary conditions, and its influence is not especially 
feared. An increase in the proportion of sulphur in cast iron 


is most likely to result from an absorption of sulphur in the 
coke during the operation of melting in the cupola. 

(4) Influence of Manganese. This element by itself de- 
creases fluidity, increases shrinkage, and makes the iron harder 
and more brittle. It combines with iron in all proportions. 
With manganese less than one-half, the combination is usually 
called spicgcleifcn. With manganese more than one-half it is 
called ferro-manganese. One of the most important properties 
of manganese in combination with iron is that it increases the 
capacity of the iron for carbon. Pure iron will only take about 
3*X per cent of carbon, while with the addition of manganese 
the proportion may rise to 6 per cent or 7 per cent. Manganese 
is also believed to decrease the capacity of iron for sulphur, and 
to this extent may be a desirable ingredient in proportions not 
exceeding I per cent to i l /2 per cent. 

(5) Influence of Chromium. This substance is rarely found 
in cast iron, but it has the property, when present in large 
proportion, of raising the capacity of the iron for carbon from 
about $y 2 per cent up to about 12 per cent. 

(6) Shrinkage of Cast Iron. At the moment of hardening, 
cast iron expands and takes a good impression of the mould. 
In the gradual cooling after setting, however, the metal con- 
tracts, so that on the whole there is a shrinkage of about % in. 
per foot in all directions, though this amount varies somewhat 
with the quality of the iron and with the form and dimensions 
of the pattern. In a general way hardness and shrinkage in- 
crease and decrease together. 

(7) Strength and Hardness of Cast Iron. The hardness of 
cast iron is chiefly dependent 'on the amount of combined car- 
bon, as noted above in (i). 

The strength is also chiefly dependent on the same ingre- 
dient. As shown in (i), the greatest crushing strength is ob- 
tained with sufficient combined carbon to make a rather hard, 
white iron, while for the maximum transverse or bending 
strength the combined carbon is somewhat less and the iron 
only moderately hard, and for the greatest tensile strength the 
combined carbon is still less and the iron rather soft. Metal 
still softer than this grade works with the greatest facility, but 
is deficient in strength. 

Numerical values for the strength will be given at a later 


(8) Uses of Cast Iron in Marine Engineering. Cast iron is 
used for cylinders, cylinder heads, liners, slide valves, valve 
chests and connections, and generally for all parts having con- 
siderable complexity of form. It is also used for columns, bed 
plates, bearing pedestals, caps, etc., though cast and forged steel 
are to some extent displacing cast iron for some of these items. 
It is also used for grate bars, furnace door frames, and minor 
boiler fittings, and for a great variety of special purposes usually 
connected with the stationary or supporting parts of machines. 

(9) Inspection of Castings. In the inspection of castings 
care must be had to note the texture of the surface, and to this 
end the outer scale and burnt sand should be carefully removed 
by the use of brushes or chipping hammer, or, if necessary, 
by pickling in dilute muriatic acid. The* flaws most liable 
to occur are blow holes and shrinkage cracks. The latter, 
however, are not often met with when the moulding and 
casting are properly carried out. The parts of the casting most 
liable to be affected by blow holes are those on the upper side 
or near the top. On this account a sinking head or extra piece 
is often cast on top, into which the gases and impurities may 
collect. This is afterward cut off, leaving the sounder metal 

The presence of blow holes, if large in size or in great 
number and near the surface, may often be determined by tap- 
ping with a hand hammer. The sound given out will serve to 
indicate to an experienced ear the probable character of the 
metal underneath. 

(10) Special Operations on Cast Iron. Cast iron may be 
softened and toughened by the process of malleablizing, as de- 
scribed in (2). It may be somewhat hardened on the surface 
by arresting the usual process of malleablizing at a suitable 
point and then hardening as for steel. This operation arrested 
before completion results in the formation of a surface layer of 
material having essentially the properties of steel. 

Cast iron may be brazed to itself or to most of the com- 
mon structural metals by the use of a brazing solder of suitable 
melting point, and with proper care in the operation. Cast 
iron may also be united to itself or to wrought iron or steel by 
the operation of burning. This consists in placing in position the 
two pieces to be united, and then allowing a stream of melted 
cast iron to flow over the surfaces to be joined, the adjacent 


parts being protected by fire clay or other suitable material. 
The result is to soften or partially melt the surfaces of the 
pieces, and by arresting the operation at the right moment they 
may be securely joined together. 

[a] Malleable Iron. 

(i) Composition and Manufacture. If a casting of hard, 
white iron, or one with a large proportion of combined carbon, 
be packed in some material which will not fuse at a red heat, 
which will exclude the air, support the piece and prevent defor- 
mation when hot, and if it be then subjected to continuous red 
heat for some days, the combined carbon will be separated from 
the iron, but will not be able to collect together in flakes or 
scales or to form the same structure as in soft, gray cast iron. 
In consequence the iron crystals remain in more intimate con- 
tact, much as in steel, and the tensile strength and toughness 
are greatly increased. 

It has long been supposed that this operation involved an 
actual withdrawal of the carbon from the iron, and to this end 
the substances usually employed are either the common red oxide 
of iron in the form of hematite iron ore, or the black oxide in 
the form of mill scale, or the corresponding oxide of manganese. 
These have a decarbonizing effect ; that is, under the conditions 
existing they will to some extent withdraw the carbon from the 
surface layer of iron. Analyses of malleable iron show, how- 
ever, that only to a slight extent is Ihe carbon actually with- 
drawn as a whole, and that the principal change is in the condi- 
tion of the carbon, as above explained. The surface effect, 
however, extending in, as it does, for perhaps 1-16 in., is un- 
doubtedly a valuable feature, and while a good quality of malle- 
able iron has been made by the use of river sand as a packing 
medium, the use of the substances mentioned above is rather to 
be preferred. 

In order that the process may be successful, the iron must 
have nearly all the carbon in the combined state, and must be 
low in sulphur, as the latter substance is found to greatly in- 
crease the time necessary. It has been customary to use only 
good charcoal-melted iron in which the sulphur is very low, 
though a coke-melted iron is quite as suitable, provided the pro- 
portion of sulphur is correspondingly small. The process can 
rarely be applied to very large castings, because such, cooling 


slowly, usually show a considerable proportion of graphitic 

To carry out the process the castings are embedded in the 
material selected. The whole is then inclosed in a cast-iron box 
or pot and is subjected to a full red heat for from two or three 
days to as many weeks, depending on the size of the piece. 

(2) Physical and Mechanical Properties. Outside of the 
numerical information, to be given later, attention may be called 
to the ductility of malleable iron, which is from four to six times 
that of cast iron, though only about one-tenth that of wrought 
iron. Nevertheless good malleable iron can be bent and twisted 
to a very considerable extent before' breaking, and its ability to 
withstand blows or shocks is very much greater than cast iron. 
Malleable iron may with care be forged and welded, and it may 
be case hardened much as with wrought iron. 

(3) Uses in Marine Engineering. Malleable iron is used 
for junction boxes and for pipe fittings in certain varieties of 
water-tube boilers, and to some extent for general pipe fittings 
on board ship. It would seem that the use of this material 
might with advantage be extended to many parts in which more 
strength and toughness are required than can be provided by 
cast iron of the ordinary type. 

[3] Wrought Iron. 

(i) Composition and Manufacture. Wrought iron is nearly 
pure iron mixed with more or less slag. Nearly all the wrought 
iron used in modern times is made by the puddling process. For 
the details of this process reference may be had to text-books 
on metallurgy. We can only note here that in a furnace some- 
what similar to the open-hearth referred to in [5] (4) most of the 
carbon, silicon and other special ingredients of cast iron are re- 
moved by the combined action of the flame and of a molten bath 
of slag or fluxing material consisting chiefly of black oxide of 
iron. As this process approaches completion small bits of near- 
ly pure iron begin to separate out from the bath of melted slag 
and unite together. This is helped along by the puddling bar, and 
after the iron has thus become separated from the liquid slag 
it is taken out, hammered or squeezed, and rolled down into bars 
or plates. Some of the slag is necessarily retained in the iron 
and by the process of manufacture is drawn out into fine threads, 


giving to the iron a stringy or fibrous appearance when nicked 
and bent over or when pulled apart. 

The proportion of carbon in wrought iron is very small, 
ranging from .02 to .20 of one per cent. In addition, small 
amounts of sulphur, phosphorus, silicon and manganese are 
usually present. 

The proportion of sulphur should not exceed .01 of one 
per cent. Excess of sulphur makes the iron red-short, that is, 
brittle when red hot. ' 

The proportion of phosphorus may vary from .05 to .25 of 
one per cent. Excess of phosphorus makes the metal cold- 
short, that is, brittle when cold. 

The proportion of silicon may vary from .05 to .30 of one 
per cent. 

The proportion of manganese may vary from .005 to .05 of 
one per cent. The influence of the silicon and manganese is 
usually slight and unimportant. 

(2) Special Properties. Wrought iron is malleable and 
ductile, and may be rolled, forged, flanged and welded. It can- 
not be hardened as steel, though by the process- of case-harden- 
ing a surface layer of steel is formed and may be hardened. 
Wrought iron may be welded, because for a considerable range 
of temperature below melting (which takes place only at a very 
high temperature indeed) the iron becomes soft and plastic, and 
two pieces pressed together in this condition unite and form on 
cooling a junction nearly as strong as the solid metal. In order 
to be thus successful, however, the iron must be heated sufficient- 
lytobringit to the plastic condition, yet not overheated, and there 
must be employed a flux (usually borax) which will unite with 
the iron oxide and other impurities at the joint, and form a thin 
liquid slag, which may be readily pressed out in the operation, 
thus allowing the clean metal faces of the iron to effect a union 
as desired. 

(3) Uses in Marine Engineering. In modern practice the 
place of wrought iron in marine engineering has been almost en- 
tirely taken by steel. Its former office was for all moving parts 
requiring strength and toughness. It is still used to some ex- 
tent for the stay bolts and braces of boilers, and for boiler tubes. 

[4] Blisters and I/aminations. 

With modern boiler material these defects are happily rare. 


In older practice, however, when wrought iron was the material 
employed for boiler plates, such defects were quite frequently 
met with. A lamination consisted in the separation of the ma- 
terial of the plate into layers not welded together and therefore 
lacking the strength and solidity of the plate proper. The for- 
mation of such places was usually due to the presence of slag in 
the iron which in the operation of rolling the plates would be- 
come thinned out into a sheet or layer .separating the two gaits 
of the iron and preventing them from becoming welded together 
and thus forming a solid homogeneous plate. Such places may 
vary in size from a trifling amount up to patches of several 
square feet in area. Now when a plate with such laminations is 
worked into the structure of a boiler with the continual fluctua- 
tions of temperature and the consequent expansions and con- 
tractions, it very frequently happens that the two parts of the 
laminations become separated from each other, and in particular 
the thinner of the two will become raised often to a considerable 
extent, thus forming a so-called blister. The chief danger from 
such a blister on the heating surface arises from the non-con- 
ductivity of the plate for heat at this point, and the consequent 
danger of its overheating and giving rise to a serious rupture. 
Further reference to this point will be found in Sec. 38 [3]. 

In examining a plate for laminations or small blisters not 
plainly shown to the eye, the hammer test is usually considered 
the most reliable. The plate is tapped over its surface, and 
judging by the sound an experienced ear can usually detect the 
locality and approximate extent of trouble of this character. 

[5] Steel. 

(i) General Composition. The properties of steel depend 
partly on the proportions of carbon and other ingredients which 
it may contain, and partly on the process of manufacture. The 
proportion of carbon is intermediate between that for wrought 
iron and for cast iron. In the so-called mild or structural steel 
the carbon is usually from i-io to 1-4 or 1-3 of one per cent. 
In spring steel the carbon proportion is somewhat greater, and 
in high carbon grades such as are used for tool steel, etc., the 
carbon is from .6 to 1.2 per cent. In addition to the carbon 
there may be sulphur, phosphorus, silicon and manganese in 
varying but very small amounts. 

From the proportion of carbon it follows that steel may be 


made either by increasing the proportion in wrought iron or 
decreasing the proportion in cast iron. The earlier processes 
followed the first method, and high-grade steels are still made 
in this way by the crucible process. 

(2) Crucible Steel. In this process a pure grade of 
wrought iron is rolled out into flat bars. These are then cut 
and piled and packed with intermediate layers of charcoal and 
subjected to a high temperature for several days. This recar- 
bonizes or adds carbon to the w r rought iron, and thus makes 
what is then called cement or blister steel. These bars, are then 
broken into pieces of convenient size, placed in small crucibles, 
melted, and cast into bars or into such forms as are desired. 

NOTE. Mild or structural steel is made wholly by the second general 
method the reduction of the proportion of carbon in cast iron. There are 
two general processes, known as the Bessemer and the Siemens Martin or 

(3) Bessemer Process. In this process the carbon and sili- 
con are burned almost entirely out of the cast iron by forcing 
an air blast through the molten iron in a vessel known as a con- 
verter. A small amount of Spiegel eisen or iron rich in carbon and 
manganese is then added in such weight as to make the propor- 
tion of carbon and manganese suitable for the charge as a 
whole. The steel thus formed is then cast into ingots or into 
such forms as may be. desired. 

In this process no sulphur or phosphorus is removed, so 
that it is necessary to use a cast iron very nearly free from these 
ingredients in order that the steel may have the properties de- 
sired. A modification by means of which the phosphorus is re- 
moved, and known as the basic Bessemer process, is used to 
some extent. In this, calcined or burnt lime is added to the 
charge just before pouring. This unites with the phosphorus, 
removes it from the steel, and brings it into the slag. In the 
basic process the lining of the converter is made of gannister or 
a calcined magnesia limestone, in order that it may not also be 
attacked by the added limestone and the resulting slag. 

In that form of Bessemer process first noted, and often 
known as the acid process in distinction from the latter or basic 
process, the lining of the converter is of ordinary fire clay or 
like material. 

The removal of the phosphorus by the basic process makes 
possible the use of an inferior grade of cast iron. At the same 


time, engineers are not altogether agreed as to the relative 
values of the two products, and many prefer steel made by the 
acid process from an iron nearly free from phosphorus at the 

(4) The Open-hearth Process. In this process a charge of 
material consisting of wrought iron, cast iron, steel scrap, and 
sometimes certain ores, is melted on the hearth of a reverbera- 
tory furnace heated by gas fuel on the Siemens Martin or re- 
generative system. The carbon is thus partially burned out in 
much the same manner as for wrought iron, and the proportion 
of carbon is brought down to the desired point or slightly be- 
low. A charge of Spiegel eisen or ferro-manganese is then 
added in order that the manganese may act on any oxide of iron 
slag which remains in the bath, and which would make the steel 
red-short if allowed to form a part of the charge. The mangan- 
ese separates the iron out from the oxide, returns it to the bath, 
while the carbon joins in with that already present, and thus 
produces the desired proportions. 

Here as with the similar operation with the Bessemer con- 
verter there is no removal of either sulphur or phosphorus, and 
only materials nearly free from these ingredients can be used 
for steel of satisfactory quality. With very low carbon, how- 
ever, a little phosphorus seems to be desirable to add strength 
to the metal. This limitation of the available materials has led, 
as with the Bessemer process, to the use of calcined limestone, 
which unites with most of the phosphorus and holds it in the 
slag. Here, as in the Bessemer process also, it is necessary to 
use a basic lining for the furnace, and it is known as the basic 
open-hearth process. By distinction the method without the 
use of the limestone has come to be known as the acid open- 
hearth process. 

As between the products of these two kinds of open-hearth 
process, there is much difference of opinion among engineers. 
Either will produce good steel with proper care, and neither will 
without it. It is usually considered sufficient to specify the al- 
lowable limits for the proportions of phosphorus and sulphur 
and leave the choice of the acid or basic processes to the maker. 

(5) Open-hearth and Bessemer Steels Compared. Open- 
hearth steel is usually preferred for structural material in mar- 
ine engineering. This is because : 

(a) It seems to be more reliable and less subject to un- 


expected or tmexplainable failure than the Bessemer product. 

(b) Analysis shows that it is much more homogeneous in 
composition than Bessemer steel, and experience shows that it 
is much more uniform in physical quality. This is due to the 
process of manufacture, which is much more favorable to a 
thorough mixing of the charge than in the Bessemer process. 

(c) The open-hearth steel may be tested from time to time 
during the operation, so that its composition may be determined 
and adjusted to fulfil specified conditions. This is not possible 
with the Bessemer process, and the latter product is therefore 
not under so good control as is the open-hearth. 

(6) Influence of Sulphur on Steel. Sulphur makes steel 
red-short or brittle when hot, and interferes with its forging and 
welding properties. Manganese tends to counteract the bad ef- 
fects of sulphur. Good crucible steel has rarely more than .01 
of one per cent. In structural steel the proportion may vary 
from .02 to .08 or .10 of one per cent. When possible it should 
be reduced to not more than .03 or .04 of one per cent. 

(7) Influence of Phosphorus on Steel. Phosphorus increases 
the tensile strength and raises the elastic limit of low carbon 
or structural steel, but at the expense of its ductility and 
toughness or ability to withstand shocks and irregularly applied 
loads. It is thus considered as a dangerous ingredient, and the 
amount allowable should be carefully specified. This is usually 
placed from .02 to .10 of one per cent. 

(8) Influence of Silicon on Steel. Silicon tends to increase 
the tensile strength and to reduce the ductility of steel. It also 
increases the soundness of ingots and castings, and by reducing 
the iron oxide tends to prevent red-shortness. The process of 
manufacture usually removes nearly all of the silicon, so that it 
is not an element likely to give trouble to the steel maker. The 
proportion allowed should not be more than from .10 to .20 of 
one per cent. 

(9) Influence of Manganese on Steel. This element is be- 
lieved to increase hardness and fluidity, and to raise the elastic 
limit and increase the tensile strength. It also removes iron 
oxide and sulphur, and tends to counteract the influence of such 
amounts of sulphur and phosphorus as may remain. It is thus 
an important factor in preventing red-shortness. The propor- 
tion needed to obtain these valuable effects is usually found be- 
tween .20 and .50 of one per cent. 


(10) Semi-steel. A metal bearing this trade name has in 
recent years attracted favorable attention among engineers and 
has come into considerable use where somewhat greater 
strength and toughness are required than can be provided by 
cast iron. 

Semi-steel is made by melting up mild steel scrap, such as 
punchings and clippings of boiler plate, with cast iron pig, in the 
proportion of about 25 or 30 per cent of the former to 75 or 70 
per cent of the latter. The presence of manganese and other 
special fluxes in small proportions is also found to add essen- 
tially to the strength, toughness and good machining qualities 
of the product. In this way is obtained a material having a 
tensile strength of 35,000 Ib. or over, and a toughness and ability 
to withstand shocks decidedly greater than for cast iron, and 
with fairly good machining qualities. Semi-steel casts as read- 
ily as most grades of cast iron, and its shrinkage and genera! 
manipulation are about the same. The chief drawback seems 
to lie in the danger of hardness under the lathe, planer or bor- 
ing tool, but with the proper mixtures this is avoided, .and a 
material very satisfactory for many purposes in marine engi- 
neering is thus produced. 

(n) Mechanical Properties of Steel. The tensile strength 
of the lowest carbon steel, say about .10 of one per cent carbon, 
is usually not above from 50,000 to 55,000 Ib. per square inch of 
section. The strength increases with the increase of carbon, 
and with not above the usual proportions o sulphur and phos- 
phorus, quite uniformly. Experiment shows that under these 
circumstances the strength will increase up to 75,000 Ib. per 
square inch, or higher, at the rate of from 1,200 to 1,500 Ib. per 
.01 of one per cent of carbon added. At the same time, with 
the increase in strength the ductility decreases, so that a proper 
choice must be made according to the particular uses for which 
the steel is intended. With the best grades of tool steel with 
carbon ranging from y 2 to i per cent and over, the strength 
ranges from 80,000 Ib. upward to 120,000 Ib., and even higher 
in exceptional cases. 

Flange and rivet steel must be tough and ductile in the high- 
est degree. Such steel has usually a tensile strength between 
50,000 and 60,000 Ib. and an elastic limit of 30,000 to 40,000 Ib. 
Its elongation in 8 in. is from 30 to 35 per cent, and reduction of 
area at the ruptured section from 50 to 60 per cent. It will 


bend cold ort itself and close down flat under hammer or press, 
up to a thickness of 1/4 in. to I in. without a sign of fracture. 

Shell steel, used for boiler shells, etc., has usually a strength 
between 55,000 and 65,000 Ib. ; elastic limit from 33,000 to 45,- 
ooo Ib. ; elongation in 8 in. of 25 to 30 per cent, and reduction of 
area of 50 to 60 per cent. 

For shafting the quality of the steel is about the same as 
for shell plates. For piston and connecting rods the strength 
is rather higher, and ductility somewhat lower. 

For steel castings the strength required is usually from 60,- 
ooo to 65,000 Ib., with an elongation in 8 in. of from 10 to 15 
per cent. 

(12) Various Specifications for Structural Steel. 

U. S. Navy. 


Phosphorus : Not over .035 of one per cent. 
Sulphur : Not over .040 of one per cent. 


Tensile Strength: Between 65,000 and 73,000 Ib. per square 

Elongation (transverse) : Not less than 22 per cent in 8 in. 

Elongation (longitudinal) : Not less than 25 per cent in 8 in. 

Elastic Limit : Not less than 35,000 Ib. per square inch. 

Cold Bending Test. One piece cut from each shell and 
curved head plate, as finished at the rolls for cold-bending test, 
must bend over flat on itself without sign of fracture. 


Tensile Strength : Between 52,000 and 60,000 Ib. per square 

Elongation : Not less than 26 per cent in 8 in. 

Quenching Test. One piece shall be cut from each furnace 
or flange plate as finished at the rolls for quenching test, and 
after heating to a dark cherry red plunged into water at a tem- 
perature of 82 deg. F. The piece thus prepared must be bent 
double round a curve of which the diameter is not more than the 
thickness of the piece tested, without showing any cracks. The 
ends of the pieces must be parallel after bending. 



Kind of Material. Steel for boiler rivets must be made by 
the open-hearth process and must not show more than .035 of 
one per cent of phosphorus, nor more than .04 of one per cent 
of sulphur, and must be of the best composition in other re- 

Tensile Tests. These specimens for rivets for use in the 
longitudinal seams of boiler shells shall have from 62,000 to 
70,000 Ib. per square inch tensile strength, with an elongation 
of not less than 25 per cent in 8 in. ; and all others shall have a 
tensile strength of from 54,000 to 62,000 Ib. per square inch, 
with an elongation of not less than 28 per cent in 8 in. 

Shearing Tests. From each heat, rivets must show a shear- 
ing strength of at least 51,000 Ib. per square inch for rivets to be 
used in longitudinal seams of boiler shells, and at least 44,000 
Ib. per square inch for all other boiler rivets. Rivets to be 
driven at the same heat used for working. 

Hammer Test. From each lot six rivets are to be taken at 
random and submitted to the following tests : 

(a) Two rivets to be flattened out cold under the hammer 
to a thickness of one-half the diameter of the part flattened with- 
out showing cracks or flaws. 

(b) Two rivets to be flattened out hot under the hammer 
to a thickness of one-third the diameter of the part flattened 
without showing cracks or flaws the heat to be the working 
heat when driven. 

(c) Two rivets to be bent cold into the form of a hook 
with parallel sides without showing cracks or flaws. 


Kind of Material. Steel for stay rods and braces must be 
made by the open-hearth process, and must not show more than 
.035 of one per cent of phosphorus, nor more than .04 of one 
per cent of sulphur, and must be of the best composition in 
other respects. 

Treatment. All material for boiler bracing must be an- 
nealed after working. 

Tensile Test. Bracing coming into contact with the fire 
must have a tensile strength of from 50,000 to 58,000 Ib., and an 
elongation of not less than 28 per cent in 8 in., or of 33 per 
cent in 2 in. in case 8-in. specimens can not be secured. Other 


bracing must have a tensile strength of not less than 65,000 lb., 
and an elongation of not less than 24 per cent in 8 in., or of 30 
per cent in 2 in. in case 8-in. specimens can not be secured. 

Bending Test. One bar l / 2 in. thick, cut from each lot of 
the bracing coming in contact with the fire, must stand bending 
double to an inner diameter of i in. after quenching in water at 
a temperature of 82 deg. F., from a dark cherry-red heat with- 
out showing cracks or flaws. A similar piece cut from each lot 
of the other bracing must stand cold bending double to an inner 
diameter of i in. without showing cracks or flaws. 

Opening and Closing Tests. Angles, T bars, etc., are to be 

subjected to the following additional tests : A piece cut from 

one bar in twenty to be opened out flat while cold ; a piece cut 

from another bar in the same lot shall be closed down on itself 

until the two sides touch without showing cracks or flaws. 


Tensile Strength : Not less than 80,000 lb. per square inch. 

Elongation : Not less than 26 per cent in 2 in. 

Elastic Limit : Not less than 50,000 lb. per square inch. 

Bending Test. One longitudinal bar y 2 in. thick, cut from 
each forging, must stand bending double, when cold, to an in- 
ner diameter of i in. without showing cracks or flaws. 


Tensile Strength : Not less than 80,000 lb. per square inch. 

Elongation : Not less than 25 per cent in 2 in. 

Elastic Limit : Not less than 50,000 lb. per square inch. 


Tensile Strength : Not less than 58,000 lb. per square inch. 

Elongation : Not less than 30 per cent in 2 in. 

Bending Test. Bars y 2 in. thick, cut from each length of 
shaft, must stand bending double to an inner diameter of i in. 
without showing cracks or flaws. 


Phosphorus : Not more than .06 of one per cent. 
Tensile Strength: Not less than 60,000 lb. per square inch. 
Elongation (for moving parts): Not less than 15 per cent 
in 8 in. 


Elongation (other castings) : Not less than 10 per cent in 
8 in. 

Bending Test. A bar I in. square shall bend cold without 
showing cracks or flaws, through an angle of 120 deg. for cast- 
ings for moving parts of machinery, and 90 deg. for other cast- 
ing, over a radius not greater than i l / 2 in. 

U. S. Inspection Requirements for Boiler Plate. 

Phosphorus : Not more than .06 of one per cent. 

Sulphur : Not more than .04 of one per cent. 

Elongation (J4 m - and under) : 25 per cent in 2 in. 

Elongation ( l /4 in. to 7-16 in. inc.) : 25 per cent in 4 in. 

Elongation (7-16 in. to I in. inc.) : 25 per cent in 8 in. 

Elongation (i in. and over) : 25 per cent in 6 in. 

Reduction of Area at Rupture (y 2 in. and under) : Not less 
than 50 per cent. 

Reduction of Area at Rupture ( l / 2 in. to 34 in.) : Not less 
than 45 per cent. 

Reduction of Area at Rupture (fy in. and over) : Not less 
than 40 per cent. 

American Boilermakers* Association Requirements. 

Phosphorus : Not over .04 per cent. 

Sulphur : Not over .03 per cent. 

Tensile Strength : 55,000 to 65,000 Ib. 

Elongation (3/ in. and under) : 20 per cent in 8 in. 

Elongation (3/ in. to 3/^ in.) : 22 per cent in 8 in. 

Elongation (34 in. and over) : 25 per cent in 8 in. 

Cold Bending. For plates y 2 in. thick and under, specimen 
must bend back on itself without fracture. For plates over y 2 
in. thick, specimen must bend 180 deg. around a mandril one 
and one-half times thickness of plate without fracture. 

British Board of Trade Requirements. 

Tensile Strength of Plates Not Exposed to Flame : 60,480 
to 71,680 Ib. per square in. 

Tensile Strength of Plates Exposed to Flame : 58,240 to 
67,200 Ib. per square inch. 

Elongation: From 18 to 25 per cent in 10 in. 


Standard Specifications Adopted by the Association of Ameri- 
can Steel Manufacturers. 

Special Open-hearth Plate and Rivet Steel. 

Steel shall be of four grades, as follows : Extra Soft, Fire- 
box, Flange or Boiler, and Boiler Rivet Steel. 

Extra Soft, Fire-box and Boiler Rivet Steel: Maximum 
phosphorus, .04 per cent ; maximum sulphur, .04 per cent. 

Flange or Boiler Steel : Maximum phosphorus, .06 per 
cent ; maximum sulphur, .04 per cent. 


Extra Soft and Boiler Rivet Steel. 

Ultimate Strength : 45,000 to 55,000 Ib. per square inch. 
Elastic Limit : Not less than one-half the ultimate strength. 
Elongation : 28 per cent. 

Cold and Quench Test : Bends 180 deg. flat on itself with- 
out fracture on outside of bent portion. 

Fire-box Steel. 

Ultimate Strength : 52,000 to 62,000 Ib. per square inch. 
Elastic Limit : Not less than one-half the ultimate strength. 
Elongation : 26 per cent. 

Cold and Quench Test : Bends 180 deg. flat on itself with- 
out fracture on outside of bent portion. 

Flange or Boiler Steel. 

Ultimate Strength : 52,000 to 62,000 Ib. per square inch. 

Elastic Limit : Not less than one-half the ultimate strength. 

Elongation : 25 per cent. 

Cold and Quench Test: Bends 180 deg. flat on itself with- 
out fracture on outside of bent portion. 

(13) Special Properties of Steel. Mild or low carbon steel 
may be welded, forged, flanged, rolled and cast. It can not be 
tempered or hardened with a proportion of carbon lower than 
about 24 f one P er cent. High carbon steel can be welded 
only imperfectly and if very high in carbon not at all. It can 
be forged with care, and cast into forms as desired. It can be 
tempered or hardened by heating to a full yellow and quenching 
in cold water or by other means, and then drawing the temper 
to the point desired. 

Mild steel should not be worked under the hammer or 
flanging press at a low or "blue" heat, as such working is found 


in many cases to leave the metal brittle and unreliable. Steel 
in order to weld satisfactorily should have a low proportion of 
sulphur, and special care is required in the operation, because 
the range of temperature through which the metal is plastic and 
fit for welding is less than with wrought iron. 

In the operation of tempering, the steel after quenching is 
very hard and brittle. In order to give to the metal the prop- 
erties desired, the temper is drawn down by heating it up to a 
certain temperature, and then quenching again, or, better still, 
allowing it to cool gradually, provided the temperature does not 
rise above the limiting value suitable for the purpose desired. 
If the reheating is done in a bath of oil the conditions may be 
kept under good control and the final cooling may be slow. If 
the reheating is in or over a fire the control is lacking and the 
piece must be quenched as soon as the proper temperature is 
reached. This is usually determined by the cplor of the oxide 
or scale which forms on a brightened surface of the metal. The 
following table shows the temperatures, corresponding colors, 
and uses for which the various tempers are suited : 

4 4 5o S$l:} Hardest and keenest cutting tool, 

470 Full yellow. ) Cutting tools requiring less hardness 

490 Brown yellow or orange, f and more toughness. 

510 Purplish. ) Tools for working softer materials, or those required 

530 Purple, f to stand rough usage. 

550 Light blue. ] Spring temper. Used for tools requiring great 

560 Full blue. [ elasticity or toughness, or for working very soft 

6oo c Dark: blue. J materials. 

(14) Special Steels. In the common grades of steel the 
valuable properties are due to the presence of carbon modified 
in some degree by other ingredients as already described. 
There are other substances which by uniting with iron in small 
proportions are able to give to the combination increased 
strength or hardness or other valuable properties. We have 
thus various special steels in which the properties may be due 
to the presence of both carbon and other ingredients, or due 
chiefly to special ingredients other than carbon. Of these 
special steels we may note the following : 

Nickel steel, containing somewhere about 3 per cent of 
nickel and varying amounts of carbon, is found to have in- 
creased strength and toughness as compared with ordinary 
steel. Nickel steel is most extensively used for armor plate, 


though to some extent it has been employed in Government 
work for screw-shafts and for boiler plates. For the former 
purposes it has given excellent satisfaction, but for the latter 
use difficulty has been met with in obtaining plates free from 
surface defects. 

Chrome steel, containing from .5 to 1.5 or 2 per cent of 
chromium may be made excessively hard, but it is not always 
reliable, and is not regarded with general favor. 

Tungsten steel or mushet steel is a steel containing carbon 
and tungsten, the latter in proportions as high as 8 to 10 per cent. 
This steel must be forged with .care and is excessively hard. 
The hardness is not increased by tempering, but is naturally 
acquired as the metal cools. Hence it is said to be self-harden- 
ing. Some specimens contain also small amounts of mangan- 
ese and silver. Its chief use is for lathe and planer or other 
cutting and shearing tools where excessive hardness is required. 

(15) Uses of Steel in Marine Construction. In modern 
practice mild or structural steel is used entirely in the con- 
struction of ships. 

The same general class of material is used for all parts of 
boilers, though the tubes are still sometimes made of wrought 

Cast steel is used for various parts of engines such as pis- 
tons, crosshead blocks, columns, bed-plates, bearing pedestals 
and caps, propeller blades, and for many small pieces and fit- 
tings. Pistons are made almost exclusively of cast steel. For 
most of the other items mentioned cast iron is still used, prob- 
ably to a larger extent than cast steel, especially where the cast- 
ings are large and complicated in form, as with columns and 

Forged, steel is used for columns, piston-rods, connecting- 
rods, crank and line shafting, and for many other smaller and 
minor parts. 

Sec. 6. I,EAD. 

Lead is a very soft, dense metal, grayish in color after ex- 
posure to the air, but of a bright silvery luster when freshly cut. 
Commercial lead often contains small amounts of iron, copper, 
silver and antimony, making it harder than the pure metal. It 
is very malleable and plastic. In engineering, lead is chiefly 
of value as an ingredient of bearing metals and other special al- 
loys. Lead piping is also used to some extent for water sue- 


tion and delivery pipes where the pressure is only moderate, and 
where the readiness with which it may be bent and fitted adapts 
it for use in contracted places. 

Sec. 7. TIN. 

Tin is a soft, white, lustrous metal with great malleability. 
Commercial tin usually contains small portions of many other 
substances, such as lead, iron, copper, arsenic, antimony and 
bismuth. It is largely used as an alloy in the various bronzes 
and other special metals. Tin resists corrosion well and in con- 
sequence is often used as a coating for condenser tubes. It is 
also used for coating iron plates, the product being the so- 
called "tin plate" of commerce. It melts at about 450 deg., 
which corresponds to a steam pressure of about 400 Ibs. per 
square inch. Due to this low melting point tin is often used as 
the composition for safety plugs in boilers. 

Sec. 8. 3INC. 

Zinc, or "spelter," as it is often called commercially, is a 
brittle and moderately hard white metal with a very crystalline 
fracture. The impurities most commonly found in zinc are iron, 
lead and arsenic. It is used chiefly as an alloy in the various 
brasses, bronzes, etc., and as a coating for iron and steel plates, 
rods, etc. The process of applying zinc for such a coating is 
called "galvanizing," and the product "galvanized" iron or steel. 
Electricity, however, is not used in the process, the articles, 
after being well cleaned, being simply dipped in a tank of melted 
zinc and then withdrawn. Slabs of zinc are also used in marine 
boilers to prevent corrosion. 

Sec. 9. AI,I,OYS. 

A mixture of two or more metals is called an t a//0;y. The 
properties of an alloy are often surprisingly different from those 
of its ingredients. The melting point is sometimes lower than 
that of any of the ingredients, while the strength, elastic limit 
and hardness are often higher than for any of them. 

Mixtures of copper and zinc are called brass. Mixtures of 
copper and tin, or copper, tin and zinc, with sometimes other sub- 
stances in small proportion, form gun metals, compositions and 
bronzes. These terms are, however, rather loosely employed. 
Various mixtures of two or more of the metals copper, tin, zinc, 
lead, antimony form the various bearing metals. 


Brass and composition are used for piping and pipe-fitting; 
globe, gate, check and safety valves ; condenser tubes and shells ; 
sleeves for tail shafts, and for a great number of small fittings 
and attachments for which the metal may be suited. The 
bronzes are employed for many of the uses of brass where more 
hardness, strength or rigidity are required. They are used with 
especial success as a material for propeller blades. 

The white metals, supported or backed by some other 
metal, such as brass, cast iron or cast steel, to give the neces- 
sary strength, are now very largely used for bearing surfaces. 


In the following proportions the numbers after the ingred- 
ients denote the number of parts in 100 of the mixture. They 
represent either the usual proportions, or the results of special 
analyses of samples, and have been collected from various 
sources. The alloys are arranged in the alphabetical order of 
their names to facilitate ready reference : 

Admiralty Bronze. Copper 87, tin 8, zinc 5. 

Aluminum Brass. Copper 63, zinc 34, aluminum 3. 

Aluminum Bronze. Copper 89 to 98, aluminum n to 2. 

Anti-Friction, A. Zinc i, iron .65, lead 78.75, antimony 

Anti-Friction, B. Copper 1.6, tin 98.13, iron trace. 

Anti-Friction, C. Copper 3.8, tin 78.4, lead 6, antimony 
1 1.8. 

Babbitt (Light'}. Copper 1.8, tin 89.3, antimony 8.9. 

Babbitt (Heavy). Copper 3.7, tin 88.9, antimony 7.4. 

Brass, Common Yellow. Copper 65.3, zinc 32.7, lead 2. 

Brazing Metal. Copper 84, zinc 16. 

Brazing Solder. Copper 50, zinc 50. 

Bush Metal. Copper 80, tin 5, zinc' 10, lead 5. 

Delta Metal. Copper 50 to 60, tin I to 2, zinc 34 to 44, iron 
2 to 4. 

Deoxidized Bronze. Copper 82, tin 12.46, zinc 3,23, iron .10, 
lead 2.14, phosphorus trace, silver .07. 

Gun Metal. Copper 89, tin 8.25, zinc 2.75. 

Magnolia. Tin ?, zinc trace, iron trace, lead 83.55, anti- 
mony 16.45. 

Manganese Bronze. Copper 88.64, tin 8.7, zinc 1.57, iron 
.72, lead .30. 


Muntz Metal. Copper 60, zinc 40. 

Navy Brass. Copper 62, tin i, zinc 37. 

Navy Composition. Copper 88, tin 10, zinc 2. 

Navy Journal Boxes. Copper 82.8, tin 13.8, zinc 3.4. 

Parsons White Metal. Copper 1.68, tin 72.9, zinc 22.9, lead 
1.68, antimony .84. 

Phosphor Bronze. Copper 90 to 92, phosphide of tin 10 to 8. 

Steam Metal. Copper 85, tin 6.5, zinc 4.5, lead 4.25. 

Tobin Bronze. Copper 59 to 61, tin i to 2, zinc 37 to 38, 
iron .1 to .2, antimony .30 to .35. 

White Metal. Lead 88, antimony 12. 


[i] Different Kinds of Tests. 

Metals may be tested for strength in various ways in ten- 
sion, by pulling apart a test piece of specified pattern and size ; in 
compression, by crushing a piece of suitable dimensions ; in cross 
breaking, by supporting a bar at two points and breaking or 
bending it in the testing machine by a load applied at an inter- 
mediate point; in torsion, by twisting apart a bar in a machine 
especially designed for the purpose ; in direct shearing, by break- 
ing a riveted or pin joint connection in the usual machine; for 
impact or shock, by letting a weight drop through a certain 
height and by its blow develop suddenly the stress in the ma- 

[a] Explanation of Terms Used. 

Ultimate Strength. The ultimate strength, of a test piece is 
the load required to produce fracture, reduced to a square inch 
of original section; or in other words, the ultimate or highest 
load divided by the original area. Thus if the area of the 
cross-section of a test piece is .42 sq. in. and the load producing 
fracture is 28,400 Ib. , the ultimate strength equals 28,- 
400 -T- .42 = 67,620 Ib. per square inch. 

Elastic Limit. The elastic limit is the smallest load, reduced 
to one square inch of area, which will produce a permanent set 
or distortion of the material. Thus in a tension test if the cross- 
section is .68 sq. in. and a permanent elongation or set is just 
produced by a load of 27,600 Ib., the elastic limit is at 27,600 -i- 
.68 == 40,600. 

Elongation. A certain length being marked off on the test 
piece as described in [3], [4], the percentage of elongation is 



found by dividing the actual extension of the Jength just before 
rupture by the original length, and reducing to per cent. Thus 
if a length of 8 in. is marked off on the test piece and if the 
length between the same marks at fracture is 10.2 in., the actual 
elongation is 2.2 in. and the percentage elongation is 220 -f- 8 = 
27.5 per cent. When a test piece is first put under load the 

f " 



Fig. 1. Test Piece for Iron Plate. 

elongation is distributed nearly uniformly over its length. This 
continues until the piece begins to neck down near the point of 
final fracture. Nearly all of the remaining elongation is re- 
stricted to the immediate vicinity of this point. Hence the per- 
centage elongation with short length of test piece may be much 


1 H g- M 

U 16'Vo 20" *J 

Fig. 2. Test Piece for Steel. 

greater than with a long piece. A few years ago, for example, 
when test pieces 2 in. long were not uncommon, the actual elon- 
gation might be nearly i in., and thus percentage elongations 
approaching 50 per cent were found. In modern practice the 
length of a test piece is usually 8 in. and values of the percent- 
age elongation over 30 per cent even with vastly superior ma- 
terial, are rarely met with. In reporting elongation the length 
used should always be stated. 


\ i-1 f\AOIOS 






r* 1 

i- *j 

II * I " * 

i- 31-0 6 * u 3T 6 * 

Fig. 3. Test Piece for Steel. 

Reduction of Area. The percentage reduction of area is 
found by substracting the final area of the section at the point of 
fracture from the original area at the same point, dividing the 


difference by the latter, and reducing to per cent. Thus if the 
original area is .68 sq. in. and the final area is .36 sq. in., the 
actual reduction is .68 .36 = .32 sq. in., and the percentage 
reduction is 3200 -=- 68 = 47.5 per cent. 

[3] Test Pieces for Iron. 

In modern practice the form of test pieces for iron is usually 

Fig. 4. Plate with Coupon. 

the same as for steel, and as described in [4] . The form of test 
piece for wrought iron plate prescribed by the U. S. Board of 
Supervising Inspectors of Steam Vessels is, however, somewhat 
different, and is illustrated in Fig. I. If the plate is 5-16 in. 
thick or less, the width at the reduced section must be one inch. 


Fig. 5. Round Test Piece. 

If the plate is over 5-16 in. in thickness, the width of the piece 
must be reduced so that the cross-sectional area at the reduced 
section shall be about .4 sq. in., but it must not be greater than 
.45 sq. in. nor less than .35 sq. in. 


Fig. 6. Bending Test. Fig. 8. Angle Test. 

[4] Test Pieces for Steel and Other Materials. 

Fig. 2 shows the form of test piece for tension prescribed 
by the Navy Department for tests of steel plate for naval uses. 

Fig. 3 shows the form prescribed by the Association of 
American Steel Manufacturers, and adopted by the U. S. Board 
of Supervising Inspectors of Steam Vessels. The test piece for 



plates is cut from a "coupon," as it is called, left on one corner 
of the plate as shown at A, Fig. 4. The U. S. law requires fur- 
ther that: 

"Every, iron or steel plate intended for the construction of 
boilers to be used on steam vessels shall be stamped by the 
manufacturer in the following manner : At the diagonal corners, 
at a distance of about 4 in. from the edges and at or near the 
center of the plate, with the name of the manufacturer, the place 
where manufactured, and the number of pounds tensile strain 
it will bear to the sectional square inch." 

Fig. 5 shows the usual round form of test piece for all ma- 
terial except plates. 

[5] Bending, Quenching and Hammer Tests. 

The nature of these tests has already been described in Sec. 
S, [51, (n), (12). 

Fig. 6 illustrates a cold bending test on a piece of steel 
plate. A drift test is also sometimes required. This is illus- 
trated in Fig. 7, and consists in driving taper drifts of con- 

Fig. 7. Drift Test. 

tinually increasing size into a punched or drilled hole until the 
diameter is increased to at least twice its original size. The 
metal must stand this test without sign of fracture about the 
edges of the hole. 

Bending tests for angle and Tee irons, as referred to in Sec. 
5, [5], (12), are also illustrated in Fig. 8. 



P 2 
J O 

88 88 

"1 CO CO VO 

88 8 

co vo O 

8 8 


7 7 

. o o 

O O "- 
M CO Tf 

to M O 
vo CO O 




M 8 0^ 


* H 




O ^O i^ 

ON m 

O O O 
t-r\ CO CO 
m ui if\ 


co >-n 

CO O 00 


E | 







O CO co 
VO t^ CO 
M M M 

rh rj- CO 

co vo "i 

t*x tx, 

O O 


i f r 

^ co ON 


^ CO Tt" 

ON vovO 

O co >- 
rh M co 

W * 

* CO CO 



M t^ OO 
t^ I s ** t"^ 

^ 3 


ON m to 

co' co' co* 
vo oo o 
oo" t^. co" 


M tN. CO 
VO ON *^ 

^ vo 

CO- > 


^^A -s 

^ A ^ 


^^ -, 





rG n 




o +* 

8 w 

O bfi,4_i t/0 

O <2JQ 

. . 

: , 


O C 

O "^ <W 

\o a G 



^\ t^ 

^5 . 8 

O \ <D x <U 
X _ , <~, \o jL 

o\ 0) 


^OQ\ ^i 

<p'o co G 8 

s o o\ *3 

O . 


h4 ^ 

I _ C/5 

1 r i G O 

. ^ (fl o ^ 

^^ 1/5 


O o c^ /^" 


O |* ^ 

O "*^ o c^ *\ 

f^ r4 ^ li ^ 

1-f ^ 


O"v0 '^ 

|J S:S 

5*1 ^1 

,' s i. i 


^^ OJ 

Ctf (1) ^) 

aJ ' ' 



4 * -4* 



1 , ' 

. ^^, 

. . . 




o o o 
o o 



8 : 









uo u^ i-r O O 

O O O O O 

O O 10 
oo vo co 

^t- -=t vo 


O O aS o ^t" 

06* * ^ 

?" 8 * 



CO 10 vo LO o 

- N Tt LOCO 

O O O 
vO -3- co 

^ S^ 

^"^ O 




G ^ : ^ 


; ; 

O *-j 5 i ^_> 



fl , cfl 

J^ ^ 







CO c3 QJ 

d ^ O 
U O 

CU ^ r^ 

S 8 o 

C ^ VH 


S : ' 

G ; ; 
'g J ; 

P G' 

' J 

FUELS. 31 



Sec. ii. COAI,. 
[i] Composition and General Properties. 

The principal fuel for engineering purposes is coal. It con- 
sists of the following chief substances. 

(A) Uncrystallized Carbon. 

(B) Volatile Hydrocarbons. Hydrocarbons are chemical 
substances formed of carbon and hydrogen in certain propor- 
tions. They often become partially oxidized* by the union of 
part of their hydrogen with oxygen, in the same proportion as 
in water. Upon the application of heat to the coal they escape 
in the form of gas, and are hence said to be volatile. 

(C) Nitrogen and Oxygen. These gases, the constituents of 
air, are also found, the latter in addition to this amount joined 
to the hydrogen as above referred to. 

(D) Sulphur. This is found in small amounts, chiefly as a 
part of the mineral known as iron pyrites-^. The proportion of 
sulphur is rarely above three per cent and usually, much less. 

(E) Ash. This consists of the earthy and incombustible 
substances present as impurities in the coal. 

Coal may be roughly divided into two chief varieties, An- 
thracite and Bituminous, with intermediate grades, Semi-anthra- 
cite and Semi-bituminous occupying the general middle ground 
between the two. In the present chapter we shall frequently 
use the terms anthracite and bituminous as denoting the general 
division into the two chief varieties, as above noted. 

Anthracite coal is sold commercially in hard, compact 
lumps, showing a shiny, smooth surface when first broken. 

* Oxidized means united with oxygen. 

t Iron pyrites is a mineral formed of iron and sulphur in the proportion 
of 46.7 parts of iron to 53.3 of sulphur. 


Bituminous coal is relatively soft and is sold commercially in 
lumps or irregular size. It crumbles easily, showing often a 
rather dull surface when broken. 

In the anthracite coal the proportion of volatile matter 
varies from 3 to 10 per cent; in semi-anthracite and semi-bit- 
uminous, from 10 to 20 per cent, and in bituminous, from 20 to 
50 per cent. The amount of ash in good coal should not ex- 
ceed from 8 to 10 pen cent, while occasionally it falls as low as 
5 per cent. Anthracite coal is graded commercially, according 
to size, the chief terms being the following, in the order of in- 
creasing size : Buckwheat, pea, chestnut, stone, egg, broken and 

[a]. Combustion. 

Combustion means simply the chemical union of a substance 
with oxygen. The oxygen is furnished by the air, which con- 
tains oxygen and nitrogen. These in air are not in chemical 
union, but simply as a mixture, in the proportion by weight of 
twenty-three parts oxygen to seventy-seven parts nitrogen in 
one hundred parts of air. When bodies enter into combustion, 
or into combustion with oxygen, heat is set free, and the prod- 
ucts formed by the combustion are very much hotter than the 
original fuel and oxygen. 

The manner in which coal burns, or enters into combustion, 
depends upon its composition and upon the nature of the fire 
and the supply of air. The elements available for the libera- 
tion of heat are the carbon and the hydrogen. Small quantities 
of sulphur are frequently present, but the amount is so small 
and the heating power so feeble that its influence may be neg- 
lected. A pound of pure carbon requires for its complete com- 
bustion, 2 2-3 pounds of oxygen, and the result is 3 2-3 pounds 
of carbonic acid or carbon dioxide in the form of gas. The total 
amount of heat set free in this operation is about 14,500 heat 
units. Now, since the proportion of oxygen in the air is about 
23 per cent, the number of pounds of air required per pound of 
carbon will be 2 2-3 -r- .23, or 2.66 -=- .23, or about 12. Similarly 
a pound of pure hydrogen requires for its complete combustion, 
8 pounds of oxygen, and the result is 9 pounds of water vapor. 
The total amount of heat set free in this operation is about 62,- 
ooo heat units. In the same way as above, it follows that the 
combustion of a pound of hydrogen will require the presence 
of 8 -f- .23, or about 35 pounds of air. The amount of hydro- 

FUELS. 33 

gen, however, is usually small, and allowing for the ash the 
amount of air necessary to barely furnish the oxygen required 
for one pound of fuel is about 12, or substantially the same as 
for one pound of carbon. In practice, however, it is found that 
this would be insufficient to maintain the draft, nor could we 
expect that the air would be so distributed as to give exactly 
the right amount of oxygen at the right place. It is, therefore, 
necessary practically, to provide a large excess of air and the 
amount actually passing into the furnaces is usually not less 
than 18 or 20 pounds per pound of coal, and may even consider- 
ably exceed this amount. At 12.5 cu. ft. per pound this will give 
for the volume of air required per pound of coal from 225 to 250 
cu. ft. and upward. 

Let us now consider the process of combustion with bitum- 
inous or semi-bituminous coal. When such coal is put on the 
fire the first result is not a combustion of the carbon, but a dis- 
tillation or driving off of the hydrocarbons in the form of gas, 
and until this operation is nearly completed there will be little 
or no combustion of the carbon. During this first operation of 
distillation, heat is absorbed by the fresh coal from the re- 
mainder of the fire for the liberation of these gases which are 
substantially the same as those forming ordinary illuminat- 
ing gas. After these gases are liberated from the coal they 
rise into the furnace and combustion chamber. Here, if 
they meet with a suitable supply of air at a proper tempera- 
ture, they will be burned, both carbon and hydrogen, and 
will thus set free all the heat which is obtainable from them. If 
the air is insufficient in amount the gases will be only partly con- 
sumed, the oxygen uniting most readily with the hydrogen and 
leaving the carbon in fine particles to form smoke or soot, accord- 
ing as they float away with the products of combustion, or be- 
come closely packed together on some of the surfaces of the 
boiler. If the air is not sufficiently hot likewise, we may have 
a partial combustion resulting in burning the hydrogen into 
water vapor and in setting the carbon free as smoke or soot as 
before. If, however, the temperature is too low the gases may 
become chilled and pass off as a w r hole unburnt, thus carrying 
away not only their own heat of combustion, but also the heat 
\vhich was absorbed for their liberation. If, on the other hand, 
hydrocarbon gases are subjected to a very high temperature be- 
fore being mixed with the air, they will become more or less 


broken up into free hydrogen and carbon in fine particles. If 
these are kept at a temperature high enough for ignition and 
are supplied with oxygen, they will burn ; but if they fall below 
the proper temperature they pass off unburnt, the carbon con- 
stituting smoke or soot, as before. Smoke is, therefore, the 
sign of a fuel containing hydrocarbons, and of a more or less 
imperfect combustion. The actual amount of fuel lost in or- 
dinary smoke is, however, quite small ; so small that it is often 
considered as having no significant influence on the question 
of economy. Hence, smoke prevention is often considered as 
hardly worth special effort, so far as the saving of fuel alone is 
concerned. There may be other losses, however, in connection 
with the general condition of which smoke is an indication, and 
any mode of design and of general operation which reduces the 
smoke formation will usually tend toward economy of combus- 

We have already seen that the conditions for burning the 
hydrocarbon gas are high temperature and an air supply above 
the grates and in the combustion chamber. We have here, 
then, one of the reasons for providing openings for the proper 
admission of air above the grates as well as underneath. 

Let us now return to the residue left on the grates after the 
escape of the hydrocarbon gases. During this part of the 
operation certain kinds of bituminous coal swell up and cake 
more or less firmly together on the grate. Such are called 
caking coals. The swelling up is due to the formation of gas 
in the midst of the coal and to its efforts to escape, while the 
caking is due to a partial' softening or melting of the substance 
under heat as the hydrocarbons are set free. Other kinds of 
bituminous coal undergo little change in their external form, 
while still others break up into small particles or grains. Those 
latter varieties are called non-caking or free-burning coals. In 
any case the residue, after the hydrocarbons are set free, is 
called coke and consists of nearly pure carbon with' ash. 

As we have already seen the carbon burns by uniting with 
oxygen, and this must take place at the burning lump itself. 
Hence, it is necessary that the air should penetrate thoroughly 
all parts of the fire, and to this end it is brought in, in part at 
least, under the grate and by the draft pressure is forced up- 
ward through the mass of burning coal. If the fire is rather 
thick the operation proceeds in the following way. The car- 

FUELS. 35 

bon and oxygen first unite in complete combustion, I pound of 
carbon tq 2 2-3 pounds of oxygen, and the product, carbon di- 
oxide, proceeds upward through the fire. As this gas comes 
in contact, however, with the cooler coal in the midst or near 
the top of the bed of fuel it absorbs some of the carbon and be- 
comes changed to a combination in the proportion of i pound 
of carbon to I 1-3 pounds of oxygen. This gas is called carbon 
monoxide. In this operation also is absorbed back again more 
than two-thirds of the heat which the first combustion had liber- 
ated. If the gas should escape unburnt, a serious loss would 
result, as only about 4,450 heat units or less than one-third of 
the heat available in the carbon would have been liberated. If, 
however, the gas finds air above the grate and a suitable tem- 
perature, the carbon which was absorbed is burnt out again and 
the corresponding heat is given back, so that the final result 
is the complete combustion of the carbon and the liberation of 
all the heat possible. The formation of carbon monoxide in 
this way shows again the need of admitting air above the grate, 
as well as underneath. This gas burns with the peculiar blue 
flame so often seen, especially after a fresh firing with anthracite 
coal, and the presence of this flame thus indicates the formation 
and recombustion of carbon monoxide in the way described. 
After the coal has all been thoroughly ignited and raised to a 
bright glowing heat, the combustion into carbon dioxide is com- 
pleted at once, and there is little or no formation of carbon 
monoxide to be burned as a gas above the grate. The thinner 
the fire the more quickly is this condition reached. 

The combustion of semi-anthracite and of anthracite coal 
proceeds in the same general manner as for bituminous coal, 
except that the period of distillation becomes shorter and less 
important as the proportion of hydrocarbon is decreased. It 
thus results that an ordinary anthracite coal burns almost en- 
tirely in the manner described for the coke residue of bitumin- 
ous coal, except that in consequence of the lower temperature 
of the fuel during the early stages of combustion, there is apt to 
be a more pronounced formation of carbon monoxide for the 
combustion of which there must be a supply of air above the 
grate as already noted. 

[3]. Impurities in Coal. Clinker Formation. 

The chief impurities in coal may be divided as follows : (A) 
Nearly infusible slate, stone, and earthy matter either in separate 


lumps or distributed through the coal as a whole, thus giving it a 
low carbon value. (B) Mineral materials more or less fusible, and 
thus capable of melting and forming a slag which uniting with 
the ash and slate forms clinker. Substances liable to be pres- 
ent in coal and which are more or less fusible at the high tem- 
peratures in the furnaces are : Potash, soda, lime and silica. The 
melting point of these substances is also considerably lowered 
by mixing with iron oxide, which is always formed by the oxida- 
tion or combustion of iron pyrites. The presence of this sub- 
stance in the coal will thus result in lowering the melting point 
of the other mineral earths and impurities, and in the greater li- 
ability to form clinker. This formation of clinker may be so 
considerable as to seriously interfere with the combustion of the 
coal, and in such cases its removal must be carefully attended to 
from time to time in order to keep the fires in good condition. 
Iron oxide, or common iron rust as we call it more famil- 
iarly, will give to the ashes a reddish tinge so that such a color 
noted in the ash may usually be accepted as an indication of the 
presence of iron pyrites in the coal, with the various results 
which have been already noted. Its presence in any consider- 
able amount is also usually shown by a yellowish or brassy ap- 
pearance of the coal. For the formation of little or no clinker 
a coal should have little or no alkali, lime, or pyrites. Such coal 
in burning gives a nearly white, soft and friable ash. 

[4]. Weathering of Coal. 

When coal is exposed to the air and weather for a con- 
siderable period of time there is a slow absorption of oxygen, 
and thus a real combustion and wasting of the fuel value 
of the coal. It thus results that the coal during this 
operation is really burning up, though at a rate so slow 
that the heat developed is hardly appreciable and the change in 
the outward appearance of the coal is so gradual as to escape 
ordinary notice. The hydrocarbons are much more readily sub- 
ject to this operation of gradual oxidation or combustion than 
pure carbon, the latter entering only with great difficulty into 
union with oxygen at ordinary temperatures. It thus follows 
that bituminous coals are much more subject to waste and 
change by weathering than anthracite coals. In addition to the 
loss due to this slow combustion there is often a gradual escape 
of gaseous hydrocarbons imprisoned within the lumps, or a 

FUELS. 37 

gradual vaporization of liquid hydrocarbons and their es- 
cape as vapor. Such losses also are, of course, more marked 
with bituminous than with anthracite coals. A bitum- 
inous caking coal often becomes changed to a non-caking coal 
after exposure to the air and weather for a considerable period 
of time. 

The chief external conditions which may influence weather- 
ing are moisture and heat. If the coal contains no iron pyrites, 
moisture is believed to slightly retard the operation of slow com- 
bustion, and thus to act beneficially rather than the reverse. If 
iron pyrites is present in the coal, the conditions are changed. 
Iron pyrites oxidize with comparative readiness at ordinary tem- 
peratures, both the sulphur and iron uniting with the oxygen. It 
thus tends to set up the operation of oxidation and to break up 
the lump into small bits, while the heat developed is a further 
aid to the continuance of the process. The oxidation of iron 
pyrites is, moreover, much aided by moisture, which, therefore, 
with such coals, becomes a distinct disadvantage. In any event 
a coal with iron pyrites may be expected to suffer more seriously 
fay weathering than one free from this substance. In extreme 
cases the oxidation of the pyrites has caused the crumbling of 
the coal into such small bits that it has become nearly worthless 
for its original purposes. 

Heat in general always increases the activity of this slow 
combustion, and hence tends' to increase the loss due to weath- 
ering. The heat developed by slow oxidation in the interior of 
large piles or masses of coal escapes with great difficulty and 
thus accumulates and raises the temperature, thus making the 
conditions still more favorable for the continuance of the 
process. So far as this effect goes, therefore, the loss would 
be more serious in large piles than in small. This is, however, 
offset by the greater difficulty which the oxygen has in penetrat- 
ing to the interior of the pile as it is larger in size. It results 
that with other things equal there is no great difference in the 
loss due to weathering with coal either in large or in small bulk. 

[5]. Spontaneous Combustion. 

We have already seen under the head of weathering that 
coal at ordinary temperatures is subject to a very slow 
oxidation, or combustion, which gradually wastes away its 
fuel value. When the coal is freshly mined this oxi- 


dation seems to be especially active due to the property 
which carbon has of absorbing or condensing gases upon 
its surface. The volume of oxygen that different coals are 
capable of absorbing varies from \ l /\ to 3 times the volume of 
the coal. The oxygen thus absorbed is very' active chemically, 
due to the fact that coming from the air it is absorbed more 
readily than the nitrogen, and is thus less diluted than in the air. 
This absorption is itself attended by the production of heat, and 
this heat, in conjunction with other conditions favorable to 
chemical action, brings about an oxidation of the hydrocarbons 
of the coal, thus generating still more heat. 

Xow, if the coal is in small bulk and well ventilated, there 
will be little chance for the gradual accumulation of the heat and 
a consequent rise of temperature. A few lumps of coal exposed 
to the open air may lose much by weathering in the course of 
six months or a year, but the heat set free will readily escape 
and the rise of temperature will be unnoticeable. If, on the 
contrary, the coal is in large bulk, or is confined in bunkers with 
little or no ventilation, the heat developed by slow oxidation 
will be imprisoned and the temperature may gradually rise to 
the point w r here active combustion will proceed according to the 
supply of air available. 

It thus appears that there may be danger from no ventila- 
tion or from insufficient ventilation. Opinions differ on these 
points, but it may probably be accepted that unless the ventila- 
tion can be made thorough, the compartment should be kept 
tight and the air excluded as much as possible. At the same 
time before such closed compartments or bunkers are entered 
with a light they should be thoroughly ventilated, especially if 
the coal is of a quality likely to freely disengage hydrocarbon 

A further important point to be noted relates to the in- 
fluence which the initial temperature has on the rapidity of 
chemical actions of this kind. Below a temperature of 100 
F. the action will go on slowly with little chance of undue heat- 
ing taking place, but as soon as the temperature rises much 
above 100 F., especially with certain coals, spontaneous ignition 
is only a question of time. 

It appears, therefore, that the true index of the danger of 
spontaneous combustion must be taken as the capacity of the 
coal for absorbing or condensing gases in its outer layers or 

FUELS. 39 

near its surface. This in turn will be shown by the amount of 
moisture which it can absorb from the air. A coal which ab- 
sorbs a large amount of moisture from the air will at the same 
time absorb a large amount of oxygen, and will, therefore, be 
relatively a dangerous coal as regards spontaneous combustion, 
while on the other hand a coal which absorbs but a small amount 
of moisture from the air will likewise absorb but little oxygen, 
and will be comparatively safe as regards spontaneous combus- 
tion. The percentage of moisture which can be absorbed from 
the air by coal is found to vary from about 2.5 to 10 per cent, 
and experience has shown that the liability to spontaneous com- 
bustion varies closely with this percentage. 

In general then the liability to spontaneous combustion de- 
pends on, 

(1) The size of the cargo or compartment, increasing as 
the bulk increases. 

(2) The size of the coal, increasing as the lumps are 
smaller, and thus present relatively more surface. 

(3) The presence of iron pyrites with moisture. Iron py- 
rites has sometimes been thought a possible direct cause of 
spontaneous combustion, but the proportion of this substance 
is small, rarely rising above 3 or 4 per cent, and the heat de- 
veloped by the combustion of sulphur and iron is very much less 
per pound than for carbon and hydrogen. The heat developed 
by the oxidation of iron pyrites is, therefore, hardly sufficient 
to do more than help along the general condition of slow com- 
bustion as referred to above. In another way, however, the 
presence of the pyrites may have an important influence on the 
result. The presence of moisture favors the oxidation of the 
pyrites and as a result of this it will swell and tend to split up 
the coal, thus decreasing the size of the lump and increasing its 
absorbing surface. It is presumably in this way that the pres- 
ence of iron pyrites tends to aid spontaneous combustion. 

(4) The quality of the coal. Bituminous and semi-bitum- 
inous coals are more liable to spontaneous combustion than an- 
thracite coal, because they are more porous and friable and 
present more absorbing surface, and, furthermore, are rich, in 
easily oxidisable hydrocarbons as already noted in [i]. In fact, 
under usual conditions anthracite coal may be considered as 
beyond danger of this character. 

(5) The amount of weathering the coal has had. If it has 


been exposed to considerable weathering and has since been 
subjected to but little breakage, then additional oxygen will be 
absorbed, but very slowly, and the danger of spontaneous com- 
bustion will be very small indeed. 

(6) The temperature of the bunkers or compartments as 
affected by the nearness to boilers, funnel, etc. 

(7) Ventilation of cargo. For ventilation to be thorough- 
ly effective cold air would have to sweep continuously and freely 
through every part, a condition hardly possible to attain with a 
coal cargo. Anything short of this may possibly increase the 
danger by supplying just about the right amount of air to create 
the maximum heating. 

The gases imprisoned within the coal, reference to which 
was made in [4] , may also escape and collect in a closed and un- 
ventilated compartment or bunker, and upon the later introduc- 
tion of a light an explosion may result. This is exactly similar 
to the way in which firedamp explosions in mines may occur. 

[6]. Corrosion. 

As a further possible result of the presence of iron py- 
rites and the sulphur which is one of its constituents, 
the corrosion of the metal surfaces of the boiler may be 
mentioned. Sulphur in burning in the presence of moisture 
may produce sulphuric acid, and this may seriously corrode such 
surfaces as it comes in contact with, especially if there is oppor- 
tunity for its gradual action during periods when the boiler is 
not in active use. The conditions necessary for the formation 
of sulphuric acid are, however, not commonly present in marine 
boilers, and the danger of corrosion from the combustion of iron 
pyrites is not usually considered as serious. 

[7], Transportation and Stowage. , 

In general anthracite coal bears transportation and handling 
better than bituminous on account of its greater hardness. It 
also stows more evenly on account of the greater uniformity in 
size of lump. 

The weight of coal in the solid lump is from 70 to 80 Ib. 
per cubic foot for bituminous grades, and from 851 or 90 to 100 
Ib. per cubic foot for anthracite grades. When broken up in or- 
dinary commercial sizes, however, its weight in bulk is usually 
from 50 to 54 Ib. per cubic foot for bituminous, and from 53 to 
58 Ib. per cubic foot for anthracite. These weights correspond 

FUELS. 41 

to an allowance of from 42 to 45 cu. ft. per ton of 2,240 Ib. for 
bituminous grades and from 39 to 42 cu. ft. per ton for anthra- 
cite grades. 

[8]. General Comparison Between Bituminous and Anthracite 


As between the two kinds of coal, bituminous burns 
more readily than anthracite, and requires a somewhat 
lower temperature for the process. This is because with 
the former the hydrocarbon gases are first driven off and burnt, 
while the coke residue is left in a light and porous condition, 
and thus well suited for intimate contact with the oxygen, and 
for rapid elevation to the temperature of ignition. On the other 
hand, with anthracite coal the hydrocarbon gases are so small 
in amount as to have little influence on the process, and the coal 
has, therefore, to burn as compact, solid lumps of carbon, with 
little opportunity for contact with the oxygen except at the 
outer surface. 

It results that under the same conditions of draft, firing, 
etc., considerably more coal can be burned per square foot of 
grate surface with bituminous than with anthracite. The excess 
of the former will depend, of course, on the special conditions, 
but will usually reach from 20 per cent to 40 per cent, or even 

From the explanation given in [2] it will be clearly seen 
that smoke and soot are chiefly the products of bituminous coal. 
With anthracite coal, immediately after firing, a slight show of 
smoke may be formed, but with neither the volume nor density 
of the smoke formed by bituminous coal, while the soot formed 
with anthracite is too small in quantity to be of any significance. 

Bituminous coal, as we have seen, is more liable to spon- 
taneous combustion than anthracite, and the loss by weathering 
is usually more serious. 

A good quality of free-burning semi-bituminous coal is 
usually considered as the best variety for all around steaming 


This fuel is made from small bits of coal, coal dust, or from 
certain grades of coal which are so soft or crumbling that they 
cannot be readily used in their natural state. The material after 
selection and removal of impurities, so far as practicable, is 


reduced to powder by grinding, and is then mixed with some 
binding material and pressed into cubes or blocks weighing 
from one to three or four pounds each. The binding material 
is usually coal-tar, asphalt, crude oil refuse, or some similar 
substance. In some cases a caking coal has been used by heat- 
ing until softening occurs and then pressing into moulds while 

The character of such fuels will, of course, depend on the 
nature of the materials of which they are made. By a proper 
choice of the ingredients or by a suitable enriching with hydro- 
carbons in the form of pitch or crude oil, a fuel of most ex- 
cellent quality may be made. The pressure to which the blocks 
are subjected is so great that the materials become closely com- 
pacted together and hold their form with no more breakage 
through handling than with a good quality of semi-bituminous 
or even semi-anthracite coal. The best grades of artificial fuel 
ring when struck and absorb little or no water, thus showing a 
compact and firm structure throughout. They ignite readily 
and burn freely without an excessive formation of smoke, hold- 
ing their shape without crumbling too rapidly on the grate. 
In evaporative power the best briquettes are the equivalent of 
good coal, from which indeed they differ chemically in no es- 
sential character. 

The weight per cubic foot and the number of cubic feet per 
ton when stowed loosely are about the same as for good semi- 
bituminous coal of like quality. If packed regularly the waste 
space is much decreased and the cubic feet per ton will be re- 
duced to from 25 to 35. 

Sec. 13. UQUID 

[i]. Composition. 

The only liquid fuel of importance to the engineer 
is either crude petroleum oil, or the residue left after 
removing from the crude oil by distillation the lighter con- 
stituents, consisting of naphtha, illuminating oil, etc. Crude 
petroleum oil is a liquid of brownish tint varying from light 
straw to almost black. It consists of a very complex mixture 
of many hydrocarbons. Some of these vaporize very easily and 
escape rapidly, even at ordinary temperatures. Such constitute 
the naphthas and gasoline. Next in order come the con- 
stituents which form common illuminating oil or kerosene. 

FUELJ. 43 

Then still heavier and denser come the lubricating oils of vari- 
ous kinds. After the removal of these there still remains a 
residue capable of yielding paraffine and vaseline, and last of all 
a certain amount of gas tar and coke. 

When the process of refining or distillation is arrested after 
the removal of the naphthas and illuminating oils, and perhaps 
some of the lubricating oil, the residue consists of a rather thick 
viscid liquid, not readily ignited as compared with the crude oil, 
but under proper conditions burning readily and with great heat- 
ing power. Such residue in the Russian oil wells on the Cas- 
pian is called astatki. It constitutes more than one-half of the 
crude oil. On the other hand, with American oils the similar 
residue is much less in amount, rarely rising to one-third of the 
crude oil. Furthermore, the processes of refining are so much 
superior in the United States that of final residue after the re- 
moval of all marketable products, there is almost nothing left. 
With the best modern methods of treatment, therefore, the use 
of the residue after partial refinement does not mean the utiliza- 
tion of a waste or by-product, but the use of a substance having 
a definite market value for other and long established uses. 
For the direct use of crude oil the same is true in still higher 
degree, so that under modern conditions the use of liquid fuel 
means simply a competition with the various other industries 
involving the use of the various products of crude petroleum oil. 

[a]. Combustion. 

For the combustion of crude oil or liquid refuse, two 
methods are in use. In the earlier and better known the chief 
essential is fhat the liquid must be "pulverized" or "at- 
omized" that is, broken up into a very fine spray and 
thus brought into intimate contact with the oxygen of the air. 
This is accomplished by special devices fitted to the furnace and 
called "pulverizers'' or "atomizers." They are of two chief 
varieties according to the means used either compressed air or 
steam. In each the oil is fed by pump or allowed to flow by 
gravity to the nozzle of the device. Here it is caught by a jet 
of air or steam, as the case may be, issuing near or through it, 
and by this means is thoroughly broken into a fine spray and 
blown into the furnace in tliis condition. Once the fire started, 
the spray is ignited as it issues from the nozzle so that the re- 
sult is a long, fiercely burning jet of flame directed into the fur- 



nace. In order to produce the conditions best for complete 
combustion, it is usually found advantageous to have fire brick 
so disposed as to take the direct action of the flame. These 
bricks become heated to a high temperature and by their radiat- 
ing action help to produce and maintain a temperature suitable 
for the complete combustion of all gaseous products formed 
from the liquid spray. In a later and on the whole more effi- 
cient method the oil is first vaporized and then introduced into 
the furnace as a vapor and there burned as such. 

Under proper conditions, and especially by the latter 
method, oil may thus be burned with little or no formation of 
smoke or soot. The absence of all soot is especially favorable 
to the maintenance of a high efficiency of operation. There is 
furthermore no clinker or ash, no cleaning of fires, no opening 
of furnace doors either for firing or cleaning, and no handling 
of ashes. 

[3], Danger of Explosion. 

The danger of explosion or of the formation of an 
explosive mixture by the slow distillation of the lighter 
hydrocarbons is considerable with crude oil unless due at- 
tention is given to the airing and ventilation of the spaces 
where such gases can collect. So long as the spaces containing 
oil are full there is no danger of any such trouble, but when they 
are partially empty the gases may collect in the vacant parts, 
forming with the air an explosive mixture which needs only a 
spark or other source of fire to explode with violence. Crude 
oil residue does not contain these lighter substances and is, 
therefore, safe from danger of this character, though in all cases 
a due attention to the matter of ventilation may be recom- 
mended. It may also be noted that the more dangerous parts 
of the oil evaporate with readiness under ordinary temperatures, 
and that crude oil exposed to the open air rapidly loses these 
constituents and becomes thereby the safer for use. This op- 
eration corresponds to the weathering of coal, and entails, of 
course, some loss, but a loss the more permissible as it makes 
the fuel the safer to use. 

[4]. Evaporative Power. 

Liquid fuel has a much higher evaporative power pound 
for pound than coal. According to chemical analysis the 
combustion of one pound of liquid fuel should liberate 

FUELS. 45 

from 20,000 to 22,000 heat units, or about one and one- 
half times as much as good coal. This, combined with bet- 
ter efficiency of operation than with coal, has given experi- 
mental results showing an evaporative power twice that of coal 
or even higher. A ratio of 1.7 : i or 1.6 : i, however, is more 
commonly considered as representing the average relation un- 
der ordinary conditions, though in some cases the advantage 
has been still less marked. 

[5]. Stowage and Handling. 

The best oil residue is of about the same density as sea 
water or slightly heavier. It will thus run from 34 to 
35 cu. ft. per ton. While, therefore, its specific gravity 
is less than that of a lump of coal, it stows much better, 
so that a given space is capable of holding from 15 to 20 
per cent more fuel in the shape of oil than in the shape 
of coal. Combining this advantage with that in evaporative 
power per pound, it follows that the final result is to very nearly 
double the capacity for steam generation per cubic foot of space 
occupied by fuel. As a further point it may be mentioned that 
oil or liquid fuel may be stowed in many places on board ship 
not available for coal or for cargo in general. Such are ballast 
tanks, double bottoms, etc. Again, the ease with which oil 
may be handled, flowing as it does by gravity and stowing itself, 
is a further point in its favor. With proper facilities a ship may 
be provided with oil much more rapidly than with coal. It 
should be added, however, that oil refuse at a low temperature 
may become quite stiff, flowing only sluggishly, especially in 
small pipes. This condition may require the provision of spe- 
cial means for heating the oil so as to insure the necessary de- 
gree of fluidity. Oil refuse is, therefore, not a fuel suitable for 
arctic exploration. 

A further advantage for liquid fuel lies in the great reduc- 
tion in the fireroom force which is possible with its use. The 
handling and firing being practically automatic, only a few men 
are required to look after the oil tanks and supply pipes in a gen- 
eral way, and one fireman can give to a large number of furnaces 
the slight general attention which they require. The chief 
work in the fireroom is, therefore, reduced to water tending, 
which remains as with coal. 


[6], Use of Oil and Coal Combined. 

In some cases the use of oil in conjunction with coal has 
given promise of good results, especially on war ships, where it 
may be of the utmost importance to be able to very rapidly in- 
crease the power developed. In such cases the coal would be 
used alone under ordinary conditions, and oil added when the 
increase of power is desired. Experiments in the Italian Navy 
show that for the most complete combustion and best efficiency 
the proportion of oil to coal should be about one of oil to five of 
coal. In the same manner as above described the oil is pulver- 
ized and blown as a spray into the furnaces, where it burns with 
the gases given off by the coal. In this way the oil furnishes a 
powerful resource for suddenly forcing the fires and increasing 
the I. H. P. developed, while the amount of oil carried or con- 
sumed is quite small compared with the supply necessary if oil 
were the only fuel used. 

[7]. Cost. 

The great drawback regarding oil fuel, and one that 
is apparently too serious to be overcome, is that relat- 
ing to its price. At ordinary figures a pound of steam would 
now cost at least as much if generated by means of oil as by 
means of coal, and if the use of oil were undertaken by several 
large steamship companies or other large consumers, its price 
would rise to a point impossible at present to foresee. The use 
of liquid fuel is quite within the reach of present engineering 
means, and may be considered as a mechanical success. Due, 
however, to the limited supply, and to the uncertainties regard- 
ing its price, its use will probably, under present conditions, be 
quite limited as a fuel for the generation of steam. 




Sec. 14. TYPES OF 

In the general sense, any receptacle in whioh steam is gen- 
erated by the application of heat is a boiler. A boiler must, 
therefore, contain three fundamental features : a place for the 
fire, a place for the water, and a division or partition between 
them. The great variety of boilers arises from the different 


Fig. 9. Scotch Boiler. 

forms which these features take, and the different manner in 
which they are arranged. The keynote of the development of 
steam boilers from the earliest forms is contained in the word 
sub-division; sub-division of the hot gases and of the water so 
that no particle of either shall be very far from the partition or 
heating surface, as it is called. If in addition to this sub-division 
provision is made for a definite flow of the hot gas along one 

4 8 


side of the heating surface and of the water along the other in 
the opposite direction, the conditions for the most efficient 
transfer of the heat of the gas through the surface into the water 
will be fulfilled. In modern boilers the principle of sub-division 
has been carried to a high degree of development, but the con- 
ditions for proper circulation are but imperfectly fulfilled. The 

o~~d^o~~6"~o~~o~~b"~o~5ToVb o c 





00'' J 

"^ ^ ' -^ *+" *** -^ \_X V-X ^ ' tDl'i W\ ^^ v ' ^^ *~* 



x< 00000000000^!: 






Reinforcing Plate 
Thick. l"Rivets 

Fig. 10. Scotch Boiler, End View. 

sub-division is obtained by the use of a large number of tubes 
or tubular elements surrounded by a shell or casing. The chief 
classification of boilers is made according to the relation of the 
water and hot gas to these tubular elements. If the gas is led 
through the inside and the water is on the outside, the arrange- 
ment is known as a fire-tube boiler. If, on the contrary, the 



gas is on the outside and the water circulates through the inside, 
the arrangement constitutes a water-tube boiler. 

Fire-tube boilers may be divided into the Return tubular or 
Scotch boiler, the Direct tubular or gunboat boiler, the Locomotive 
boiler, the Flue and return tubular or leg boiler, and the Flue boiler' 

Fig. 11. Scotch Boiler, Longitudinal Section. 

as used on western river steamers. These are illustrated in 
Figs. 9-14. 

Water-tube boilers are found in great variety, depending 
on the details of arrangement of the tubes, and drums or head- 
ers of which they are composed. A few representative types 
are shown in Figs. 15-26. We will first give brief descriptions 


of the important features of these boilers, and then take up at a 
later point the subjects of their design and construction. 

[i] The Scotch Boiler. 

In present practice, for marine purposes, the Scotch boiler 
is used more than any other one type, and, in fact, more than 
all other types combined. This boiler, as illustrated in Figs. 9, 
10, ii consists essentially of a cylindrical shell containing one or 
more cylindrical furnaces, usually corrugated circumferentially 
for strength, opening into combustion chambers at the back end, 
from which a large number of small tubes lead again to the front 
end or head of the boiler. The grates are placed at about the 
center of the height of the furnace, and the fire and hot gases 
occupy the upper part of the furnaces, the combustion cham- 
bers, and the inside of the tubes, while the water and steam 
fill all the remaining parts of the shell, the water level being 
usually some 6 in. to 8 in. above the highest part of the tubes or 
combustion chambers. The hot gases pass from the fire on the 
grate-bars into the combustion chamber, thence forward 
through the tubes and out through the uptake or front-connection 
to the smoke-stack or funnel. Several varieties of this boiler are 
in common use. Thus the number of furnaces may be one, two, 
three, or four. They may be fitted with separate combustion 
chambers, or there may be one combustion chamber for all fur- 
naces, or, as is common with four furnaces, there may be two 
combustion chambers one for the two furnaces on either side. 
Again, the boilers may be single-end or double-end. Fig. 11 is 
an example of the first. A double-end boiler consists of two 
sets of furnaces opening from either end of a shell of double 
length. It is evidently equivalent to a pair of single-end boilers 
placed back to back with the back heads removed and the shells 
joined. Such boilers may also have either separate combustion 
chambers for each end, or a common combustion chamber for 
both ends. The former arrangement is to be preferred, and 

becomes necessary where forced draft is used. 

[2] Direct Tubular Boiler, Gunboat Type. 

This boiler is rarely used except in war ship practice, where 
with low head room it has been occasionally employed. It con- 
sists of a shell with furnaces and combustion chamber some- 
what as in the Scotch boiler, but the tubes, instead of returning 
to the front, fead on to the farther head. To this head is fitted 


a smoke-box or uptake leading to the funnel. In such cases 
the boiler for the same power is of smaller diameter and greater 
length than the Scotch type, and it is readily seen that the whole 
arrangement is simply a mode of exchanging diameter for 

[3] Direct Tubular Boiler, I/ocomotive Type. 

The locomotive type of marine boiler as illustrated in Fig. 
12 consists of a cylindrical shell extended to the front and modi- 

Fig. 12. Locomotive Type Boiler. 

fied in form with flat sides and bottom, and flat or rounded top. 
The furnace is of rectangular cross-section, and is surrounded 
by the shell at the front, leaving on the sides a narrow space 
known as the water-leg and sometimes a like space underneath 
known as the water bottom. The gases take the same general 
course as in the gunboat type, the chief difference in the two 
being in the form of the furnaces and in the absence of the 
combustion chamber in the locomotive type. 

[4] The Flue and Return Tubular or I/eg Boiler. 

In this boiler, as illustrated in Fig. 13, the hot gases pass 
from the furnace through large tubes or Hues, as they are 
termed, to a combustion chamber at the farther end. They 
then return to the front through small tubes, and are led by an 
appropriate uptake to the funnel. The furnace is of rec- 
tangular cross-section, and the front end of the boiler is modi- 
fied on the sides and bottom to correspond to this form, as in 
the locomotive type. Water legs are also formed in the same 
way on the sides of the furnace, and from this feature the boiler 
receives its common name. This form of the front end of the 
boiler with flat sides and rounded top is sometimes known as a 
wagon-top. Very commonly, as shown in Fig. 13, an attachment 
to the shell, known as a steam chimney, surrounds the lower part 


of the funnel, the office of which is to subject the steam to the 
drying and superheating effects of the gases on their way 

Fig. 13. Leg Boiler. 

through the funnel. Boilers of this type have been used to a 
considerable extent on tug and river boats. 

[5] The Flue Boiler. 

In Western River practice use is quite commonly made of 
the return flue boiler as illustrated in Fig. 14. This boiler 
is externally fired. The flames and hot gases pass back along 
the outside of the boiler to a back connection and then enter 
the flues and return through them to the front, and thence to 
the uptake and funnel. Boilers of the locomotive type, or 
tubular fire-box boilers as they are often called, are also used to* 
a considerable extent in western river practice. 



[6] Water Tube Boilers. 

Turning now to this type, a brief description will be given 
of the leading features, which may be combined in the greatest 
possible variety, thus giving the vast number of forms of such 
boilers on the market at the present time. To aid in the descrip- 

Fig. 14. Return Flue Boiler. 

tion a few typical forms of such boilers are shown in Figs. 15-26. 
Most boilers of this type have one or more cylindrical 
drums or chambers on top and one or more similar drums be- 
low, the two sets of drums being connected by sets of tubes. 
The feed usually enters first the upper drum, frequently passing 
on its way through a coil heater in the base of the stack or top 
of the boiler. It then flows down certain of the tubes to the 
lower drums. If these tubes are of extra large size and specially 
intended for down flow, the boiler is said to have special down 
flow or down cast tubes or pipes, as shown in Figs. 15, 16, 17, 18, 
20. In some cases such tubes are omitted, and the feed must 
descend through part of the small inner tubes. In any case, 
after finding its way to the lower drum it enters the up How or 
steam forming tubes, which are surrounded by the hot gases 
coming from the furnace below them. During the passage of 
the water upward it is partly converted into steam, and the 
mixture issues from the upper end of the tubes into the upper 
drum. There the steam is separated and led to the engine, 
while the water joins that already in this drum, and thus begins 
another round. In some cases the upper ends of the steam 
forming or delivery tubes are below the level of the water in 
the upper drum, and they are then said to be drowned or wet. 
In other cases they are above the water level and are said to 
be dry. In still other forms they enter at about the middle of the 
drum or about the water level, and may be wet or dry as the 
level varies. See the various cuts for examples. Water tube 
boilers are often divided into two general classes : large tube and 






small tube boilers. In the former they are usually 3 or 4 or 
even 5 in. dia., while in the latter they are usually from ij4 to 
\y 2 or 2 in. dia. 

Again, the tubular elements may be made up in a great 
variety of forms. In some they are straight, in others curved, 
as shown in the various figures. In small tube boilers they are 
very commonly curved or bent, while in the large tube types 
they are straight. Also in some types the elements are con- 
tinuous between drums or headers as in Figs. 17, 18, 19, 20, 
while in others, as in Figs. 15, 16, 26, they are made up of 
lengths or of different parts with screwed joints, elbows, re- 
turns, junction boxes, etc. In some they are expanded into the 
shells of the drums ; in others screwed. In some all joints are 
carefully protected from the direct action of the flame ; in 
others screwed joints are freely exposed to the flame. In some 
the general direction of the tubes is nearly horizontal ; in others 
nearly vertical, and in others bent or curved in various forms. 
In some types, as illustrated in Figs. 24, 25, the lower drums are 
omitted, or consist merely of the lower portions of the tubes 
and headers, or members to which the tubes are connected. In 
all cases the grate lies below the tubes and frequently between 
the lower drums, as shown in the various figures, while the 
whole is surrounded by a casing intended to prevent, so far as 
possible, the loss of heat by radiation. 

[7] Relative Advantages of Different Types of Boilers. 

For large ships under ordinary conditions and where the 
extremes of lightness or of speed on a given displacement have 
not to be attained, the Scotch boiler seems at present to be 
considered as fulfilling most satisfactorily the all around re- 
quirements for a marine boiler, and in consequence it is found 
almost universally in the mercantile deep sea marine, as well as 
on the Great Lakes, and to a large extent on inland craft of all 
descriptions, except those of small size. It is also used to some 
extent in naval practice, though the use of water-tube boilers 
is at the present time extending quite rapidly into this field, 
where their special features become of marked value. The pres- 
ent is a time of change with regard to types of boiler. It is 
not too much to say that in most of the modern naval construc- 
tion the water tube boiler is accepted as the standard type to 
the exclusion of the fire-tube boiler. The water-tube boiler 
is also making large advances in the mercantile field, and not a 








few modern ships of the mercantile marine are now equipped 
with this type of boiler. 

For tugboats, river steamers, and a variety of small craft, 
the various types of direct fire-tube and flue boilers have been 
much used. These boilers are more readily adapted to a var- 
iety of demands regarding size, form and arrangement, and in 
small sizes are perhaps more cheaply built than Scotch boilers. 
In many cases, however, the preference for boilers of this type 
has doubtless depended on local and special conditions quite in- 
dependently of their relative value from the engineering stand- 

For fast yachts, launches, all craft of the torpedo-boat type, 
and in fact in all cases where the highest speed is to be attained 
on the least weight, the water-tube boiler has become a neces- 
sity, and in one or another of its many forms is universally em- 

The weight of Scotch boilers without water, per square 
foot of heating surface, is usually from about 25 to 30 Ib. ; of 
water-tube boilers of the lighter types from 12 to 20 Ib. The 
weight of the contained water per square foot of heating surface 
is usually from 12 to 15 Ib. for Scotch boilers, and from, say 
1.5 to 3 Ib. for water-tube boilers. It results that Scotch boil- 
ers with water will weigh from, say, 35 to 50 Ib. per square foot 
of heating surface, while water-tube boilers will similarly weigh 
from 13.5 to 23 Ib. These figures are not to be considered as 
giving absolute limits, but simply as representative values for 
average types. It should be noted, however, that a square foot 
of heating surface in a Scotch boiler seems to be somewhat 
more efficient than in a water-tube boiler. It is difficult to es- 
timate the difference numerically, but other conditions being 
equal, it would probably be safe to give to the water-tube boiler 
additional heating surface to the extent of from ten to twenty 
per cent. On the other hand, it must be remembered that 
water-tube boilers can stand forcing to a much higher degree 
than fire-tube boilers. With the latter supplying steam to triple 
expansion engines the ratio of heating surface to I. H. P. can 
hardly be recluced below 2, while with the former this ratio has 
been reduced in many cases to less than one and one-half, and, 
as reported in certain extreme cases, to between one-half and 
one. Water-tube boilers have the further advantage that they 
are more readily constructed for the higher and higher steam 






pressure which modern practice is continually demanding. 
With water-tube boilers, due to the construction and to the 
smaller amount of contained water, there is also less danger 
from disastrous explosion. With water-tube boilers steam may 
be raised much more quickly than with fire-tube boilers ; from 
one-quarter to one-half hour is sufficient with the former, while 
from three to four hours should be taken with the latter. 

Water-tube boilers are also much more portable than fire- 
tube. In many forms spare parts or even the whole boiler may 
be shipped in elements or sections across country by rail or to 

Fig. 23. Lagrafel and D'Allest Boiler. 

foreign ports by ship transport, put on board the steamer for 
which they are intended, and erected in place without difficulty. 
On the other hand, the water-tube boiler imperatively re- 
quires fresh water feed. Under modern conditions this should 
be provided no matter what the type of boiler in use, but if in 
emergency salt water must be used, the fire-tube boiler will re- 
ceive the lesser injury. Again, from the small amount of water 
contained as a stock upon which to draw, the water-tube boiler 
requires a more uniform feed than the fire-tube boiler, and is 
generally more sensitive to variations in the conditions under 
which it works. Again, the rupture of a tube is a more serious 


Fig. 24. Babcock and Wilcox Boiler. 


<>0 55 50 45 40 

Fig. 25. Arrangement of Fireroom with Babcock and Wilcox Boilers. 






Fig. 26. Belleville Boiler. 


matter in the water-tube than in the fire-tube boiler. In the 
latter it may be plugged without disturbing the water and steam 
in the boiler, and with only a momentary interruption to its 
operation. In the former it is usually necessary to disconnect 
the boiler, draw the fires, blow down the water, and plug or in- 
sert a new tube. Water-tube boilers are also not readily made 
in large sizes or units. Scotch boilers may be made in 2,000 
horse power units or even larger, while half of this or less is 
about the maximum for the water-tube boiler. An outfit of the 
latter for large power requires, therefore, a large number of 
boilers with a corresponding increase in the fittings and attach- 
ments. On the other hand, the temporary removal of one 
boiler for repair is of less importance, as the size is decreased 
and number increased. 

To summarize the general comparison between water-tube 
and fire-tube boilers, the former have relative advantages in the 
following chief points : Weight, ability to stand forcing, suit- 
ability for high pressures, greater safety from disastrous explo- 
sion, and quickness of raising steam. On the other hand they 
have relative disadvantages in these points : A more rigid re- 
striction of the feed to fresh water, the necessity of greater 
regularity of feed, greater difficulty in dealing with leaky tubes, 
and general sensitiveness to variation in the conditions of use, 
to which may be added the present feeling of uncertainty as to 
their durability and efficiency under the conditions prevailing on 
deep water voyages. 


The various joints in a boiler are usually of the riveted 
form. The use of welded joints in various parts of boiler con- 
struction is increasing somewhat as greater skill is acquired in 
making them, but in ordinary practice the joints are riveted, and 
of various types, as follows : 

Riveted joints are divided into lap joints and butt joints, 
according as the plates lap over each other (see Figs. 31-34), 
or butt together at the edges, and are covered by one or two 
butt straps (see Figs. 35-39). They are also divided according 
to the number of rows of rivets into single, double or triple 
riveted joints (see Figs. 31-33). 

The rivets are usually staggered in arrangement, as shown 
in Figs. 32-39. Sometimes, though rarely, the chain arrange- 



ment, as shown in Fig. 27, is used. While chain riveting is as 
strong, or ( perhaps even slightly stronger, than staggered rivet- 
ing, the latter gives a better disposition of rivets for making a 
steam or water-tight joint, and this fact leads to its more fre- 
quent use in boiler construction. In butt joints the arrange- 
ment of rivets is duplicated on each side of the joint, and the 
style of riveting is named according to the arrangement on one 
side. Thus, Fig. 35 shows a double-riveted and Fig. 36 a triple- 
riveted butt joint. 

A riveted joint may fail : (i) In the plate by tearing out or 
across from hole to hole, see Figs, 28, 29; (2) In the rivet by 
shearing; (3) in the plate or rivet by a crushing of the material. 

Fig. 27. 

The failure of a joint by the tearing out of the plate in front 
of the rivet, as in Fig. 29, is safely guarded against by placing 
the row of rivets at a proper distance from the edge of the plate. 
This, by experience, is found to be about one diameter in the 
clear, or one and one-half diameters from edge of plate to center 
line of rivets. In lap joints and butt joints with one cover, as 
in Fig. 30, the rivets resist shearing at one section only. In 
butt joints with double covers, as in Figs. 35-38, the rivets re- 
sist shearing at two sections. The total shearing strength of a 
rivet in double shear is usually taken as somewhat less than 
twice the strength in single shear. The British Board of Trade 
rules give ij4 as the ratio to be used. 

With usual proportions the last mode of failure mentioned 
above is the least likely to occur, so that so long as the proper 
limits are not exceeded the resistance to crushing needs no es- 
pecial examination. These limits will be given in detail at a 
later point. 

The strength of a riveted joint is, of course, determined by 



whichever is the weaker of the two, the plate or the rivets. In 
a properly designed joint the strength of the plate and that of 
the rivets should be equal, so that there will be no more likeli- 
hood of failure in one way than the other. It may be remarked, 
however, that since corrosion usually affects the plate only, it is 
often considered good practice to give to the plate a slight ex- 

Fig. 28. 

Fig. 29. 

cess of strength, so that even after some wasting by corrosion 
the joint may still be in fair proportion as to the relative 
strength of plate and rivets. No exact directions can be given 
for this increase, as it is simply a matter of judgment. 

The investigation of the strength of riveted joints by any 
simple theory is necessarily quite imperfect, because we do not 
know in just what way the stress is distributed through the re- 
maining part of the plate, nor through the section of the rivet, 

Fig. 30. 

nor what allowance to make for the frictional grip of the joint. 
The proportions given by the following equations, however, are 
those which will give practically equal strength of plate and 
rivets, using the British Board of Trade rules. These rules 
represent standard and reliable practice, based on wide experi- 
ence, and are substantially adopted by the United States inspec- 
tion authorities. In thus considering a joint we take simply 
an element such as that between AB and CD in the following 
diagrams. It is clear in each case that the whole joint may 
be considered as made up of a series of such elements : 


Let p denote the pitc h of the rivets : that is, the distance from 

center to center. Where the rivets in one row are 

pitched twice as far apart as in another (see joints 

D, H, etc.), p denotes the larger of the two values. 

" d denote the diameter of rivet. * 

" n =: p -r- d = number of rivet diameters in the pitch. 
" t = thickness of plate. 
a = t + d. 

" T = tensile strength of plate per square inch of section. 
" S = shearing strength of rivet per square inch of sec- 

The ratio of 5 to T is taken as 23 : 28 or 5" = .821 T. 
The efficiency of the joint is the ratio between the strength 
of the joint and the original strength of the plate. It will be 
seen by the formulae given later that the efficiency of a joint is 
increased as d and p are made larger. There is, however, a 
practical limit to the increase in d, due to the difficulty of head- 
ing up very large rivets, and a limit to the increase in p, due to 
the necessity of guarding against leakage. If the general pro- 
portions between d and /, as indicated later in connection with 
the various joints, are observed, the result will be a pitch within 
safe limits, and a joint agreeing well with the best practice. 

The largest permissible values of the pitch, according to 
the Board of Trade rules, are given by the following formula : 

p = Ct + if 
where C is drawn from the following table : 






I "*! 




2 62 




7 47 


6 oo 


4. 14 


6 oo 


i CQ 

In no case should the pitch exceed 10 inches. 
We will now proceed with the equations and proportions 
for various forms of riveted joints. 

* Strictly speaking the diameter of the rivet hole should be used, as it is 
about 1-16 inch larger than the rivet before heading up. In the Board of 
Trade Rules, however, the diameter of rivet is used. The difference in 
proportion of joint is quite small, and probably not of practical importance. 

Joint A. 

Lap Joint. 

Single Riveted. 





P i 

Fig. Z 


1. Joint 

D ' 


B 3 d 

The element is A B D C, containing one rivet. We have in 
this case : 

Strength of Plate = t (p d) T 

Strength of Rivet = $xd 2 S 

For equal strength of plate and rivet, 

--or;/= i + -645-7 

Efficiency = 

n i 


a -f- .645 

The ratio d -r- t may vary from 1.5 to 2.5, the lower values 
being more commonly employed with very thick* plates on ac- 
count of the difficulty of heading up excessively large rivets, 
and the necessity of a moderate pitch to insure against leakage. 
In order, furthermore, to guard against danger of rupture by 
crushing, the upper limit, 2.5, should not be exceeded. 

The foregoing operations may be expressed also by the 
following : 

Rule, (i) Select a diameter of rivet according to the thick- 
ness of the plate and the directions given. 

(2) Multiply this diameter by .645 and divide by the 

thickness of plate. 

(3) Add i to the result obtained in (2). 

(4) Multiply the diameter of rivet by the result ob- 

tained in (3), and the result will be the pitch 
suited to the diameter chosen. 


(5) Select the nearest working dimension, going 

usually above in order to give slight excess of 
strength to the plate. 

(6) To find strength of plate in the joint, subtract the 

diameter of rivet from the pitch, multiply by 
the thickness and by the tensile strength per 
square inch of section. 

(7) To find strength of rivet, find area of section, 

multiply by the same strength per square inch 
as in (6), and then by .821. 

(8) To find original strength of plate multiply pitch 

by thickness of plate, and by the tensile strength 
per square inch, as in (6). 

(9) To find the efficiency, divide the lower of the two 
results found in (6) and (7) by that found in (8). 

Example. To lay out a single riveted lap joint for J/ inch 
plates, using % inch rivets. 

Then 7/e X .645 -*-*== 7 X '^ X ' = 1.13 


And 1.13 + i.oo = 2.13. 

And 2.13 X J = 1.86 = pitch. 

Take the nearest eighth above and we have pitch = i% 

Then taking strength of plate at 60,000 Ibs. per square inch 
we have : 

Strength of Plate in Joint =(ij J) X -J X 6,000 = 30,000. 
Area of inch rivet = .60 sq. in. 
Strength of rivet = .60 X 60,000 X -82! = 29,550. 
Original strength of plate = if X -J X 60,000 = 56,250. 
Efficiency = 29,550 -f- 56,250 = .525. 

Similarly, if we should take 15-16 inch rivets, we should find 
for equal strength in plate and rivet a pitch of 2.07 inches. If 
we take a pitch of 2 T / 8 inches we shall find an efficiency of .533. 
If we take the 2.07 exact the efficiency will be .548. 

Joint B. 


Lap Joint. 

Double Riveted. 


Fig. 32. Joint B. 

q not less than (. 6 / -|- . 4 d~} 
Hence H not less than y (i. i / -1- .4 d) (. i / -f .4 d) 

Where there are two or more rows of rivets they must be 
placed at a sufficient distance apart, so that there may be no 
danger of rupture along a zig-zag line, as indicated in the di- 
agram. To this end the British Board of Trade rules give cer- 
tain values for the distance q as given above for this case. This 
distance is known as the diagonal pitch. The rules are derived 
from experiment. The distances resulting may be considered 
as the smallest allowable. In practice the values of q are often 
made somewhat greater than would result from the rules. 
Having selected the distance q, the location of the second row 
of rivets is easily found from the first by constructing a triangle 
with base equal to />, and the two other sides each equal to q. 

In this case the element A B D C contains one whole rivet 
and two halves, or two rivets in all. We have then : 
Strength of Plate t (p d) T 
Strength of Rivets = y*7td*S 

For equal strength of plate and rivets : 

- -,- or n = i + i. 29 

Efficiency = 

n i 


a -f- 1.29 

The values of d -.- t may vary through about the same 
range as in joint A, above, and for the same reasons as there 


explained. These operations may be expressed by a rule 
similar to that for joint A, the numbered sections referring to 
that rule as given above : 

(1) Same as for joint A. 

(2) Use 1.29 instead of .645. 

(3), (4), (5), (6) Same as for joint A. 

(7) Take twice the strength of one rivet, found as for 
joint A. 

(8), (9) Same as for joint A. 

Example. To lay out a double riveted lap-joint for y 2 inch 
plates, using J4 inch rivets. 

Then % X 1.29 *- % = 3 X 1 '' 9X 2 = 1-935 


And 1.935 + LOO = 2.935. 

And 2.935 X f = 2.201 = pitch. 

Taking the nearest eighth above we have p = 2^4 inches. 

Then taking strength of plate at 60,000, we have : 

Strength of plate in joint = (2j - - f ) X i X 60,000 = 

Area of J4 mcn rivet = .4418 sq. in. 

Strength of rivets = .4418 X 2 X 60,000 X -821 = 43,520. 

Original strength of plate = 2.\ X 4 X 60,000 = 67,500. 

Efficiency = 43,520 -f- 67,500 = .645. 

Similarly with % inch rivets, pitched 2% inches, the 
strength of plate and rivets will be nearly equal, and the effi- 
ciency will rise to .687. 

Joint C. Lap Joint. Triple Riveted. 

h = a d 

q not less than (.6/ -}- .4^) 

Hence //"not less than y (i. i / -f .4 </) (. i /> -f .4 <^) 
B = 3 rf + 2 // 

In this case the element A B D C contains two whole rivets 
and two halves, or three rivets in all. We have then : 

Strength of Plate t (p d) T 

Strength of Rivets %nd*S 
For equal strength of plate and rivets : 


P d 

-or n = i -r- 1.935 


"R flfi r* i i^n r\7 


^_ AC 




11 Q' 


i C 



--^7- ^ 

B r 

H X Q v /\ 





i U 



;.__ p ^ 


Fig. 33. Joint C. 

The values of d -r- t may vary through about the same 
range as in joint A, above, and for the same reasons as there 
explained. These operations may be expressed by a rule sim- 
ilar to that for joint A, the numbered sections referring to that 
rule as given above. 
Rule : 

(1) Same as for joint A. 

(2) Use 1.935 instead of .645. 

(3), (4), (5), (6) Same as for joint A. 

(7) Take three times the strength of one rivet found as for 
joint A. 

(8), (9) Same as for joint A. 

Example. To lay out a triple-riveted lap-joint for y 2 inch 
plates, using 1/4 inch rivets. 

3 X 1-935 X 2 

Then ^ X i-935 


And 2.903 -h i. oo = 3.903. 

And 3.903 X f = 2.93 inches = pitch. 

Take p = 3 inches. 

Then taking strength of plate at 60,000, we have : 

Strength of plate in joint = (3 3) X i X 60,000 = 67,500. 

Area of ?4 inch rivet = .4418. 

Strength of rivets = .4418 X 3 X 60,000 X -821 == 65,280. 

Original strength of plate 3 X X 60,000 = 90,000. 

Efficiency = 65,280 -f- 90,000 .725. 


Similarly with % inch rivets, pitched 3% inches, the effi- 
ciency becomes .764. 

Joint D. Lap Joint, 

inner row spaced one-half p. 

Triple Riveted, with rivets in 




"I" 1 





Fig. 34. Joint D. 

h = 

q not less than (.$/ + d) 
Hence H not less than -\/ (. 5 5 / -f d} (. 05 p -\- d) 

B = 3 d + 2 H 

As seen below, the efficiency of this joint is superior to that 
of joint c, but it is perhaps slightly inferior as regards tightness 
against leakage. We have in this case : 

Strength of Plate t (p d) T 
Strength of Rivets = xd 2 S 
For equal strength of plate and rivets : 


or n 

Efficiency = 

n i 


a + 2.58 

The values of d -T- t may vary through about the same 
range as in joint A above, and for the same reasons as there 
explained. These operations may be expressed by a rule sim- 
ilar to that for joint A, the numbered sections referring to that 
rule as given above. 
Rule : 

(1) Same as for joint A. 

(2) Use 2.58 instead of .645. 

(3)i (4), (5) ( 6 ) Same as for joint A. 


(7) Take four times the strength of one rivet found as for 
joint A. 

(8), (9) Same as for joint A. 

Example. To lay out a triple-riveted joint as in D for l / 2 
inch plates, using % inch rivets. 

Then % X 2.58 % = 3J< 2 -5* X 2 = ^ 


And 3.87 + i. oo = 4.87. 

And 4.87 X t = 3.65 inches = pitch. 

Take p = 3 11-16 inches. 

Then /> = i 27-32 inches. 

Then taking strength of plate at 60,000 we have : 

Strength of plate in joint = (3---- J) X i X 60,000 = 

Area of f inch rivet = .4418. 

Strength of rivets = .4418 X 4 X 60,000 X -821 =. 87,050. 

Original strength of plate 3--X i X 60,000 = 110,625. 

Efficiency = 87,050 -r- 110,625 = .787. 

With % inch rivets spaced 2 7-16 inches in the middle row 
and 4% in the outer rows, the strength of plate and rivets 
would be nearly equal, and the efficiency would rise to .81. 

Joint E. Double Butt-Straps. Double Riveted, 
h = ij d 
q not less than (.6 p -f- .4 d) 

Hence //"not less than j/ (i. i / 4. .4 d) (. i / -j- .4 d) 
B = 6 d + 2 H. 

Thickness of each butt-strap not less than ^ the thickness 
of plate. 

The arrangement of rivets is duplicated on either side of 
the joint line P Q. We need only to investigate the part of the 
joint on one side of P Q. The element is then A B D C, as in 
joint B, except that the rivets are in double shear instead of 
single shear. For the total shearing strength of a rivet in 
double shear, as previously explained, it is customary to take 
1 24 times the strength for single shear instead of 2 times, or to 
take the two strengths in the ratio 7 : 4. 

We then have : 

Strength of Plate t (p d) T 
Strength of Rh'cts = Ind'S 

For equal strength of plate and rivets : 

-^-or ;/ = i 4- 2.26- 
d t 

n i 2.26 






Fig. 35. Joint E. 

In all double butt-strap joints d -i- / usually varies from i 
to ij4- The lower range of values, as compared with joints in 
which the rivets are in single shear, is required in order to in- 
sure the joint against danger of failure by crushing. 

These operations may be expressed by a rule similar to that 
for joint A, the numbered sections referring to that rule as 
given above. 
Rule : 

(1) Same as for joint A. 

(2) Use 2.26 instead of .645. 

(3)> (4) (5) (6) Same as for joint A. 

(7) Take 35^ times the strength of one rivet, as found for 
joint A. 

(8), (9) Same as for joint A. 

Example. To lay out a joint as in E for I inch plates, using 
ij/s inch rivets. 

Then ij X 2.26 -5- i = 9X2 ' 26 = 2.54. 


And 2.54 + i = 3.54. 

And 3.54 X-f-== 3.98 inches = pitch. 

Take pitch = 4 inches. 



Then taking strength of plate at 60,000 Ibs., as before, we 

Strength of plate in joint = (4 - - ij) X I X 60,000 

Area of i-J inch rivet -994- 

Strength of rivets = .994 X 3i X 60,000 X ,821 = 171,350. 

Original strength of plate = 4 X i X 60,000 = 240,000. 

Efficiency = 171,350 -=- 240,000 = .714. 

Similarly with i inch plates and \ l /\ inch rivets, pitched 
4^4 inches, the strength of plate and rivets will be about the 
same, and the efficiency is .737. 

oint F. Double Butt-Straps. 7 

>/>/< R 




n n n 





^ ^ ^ ^ 

_A__ ..C 






q' H 






^g . -^ .^. ^ ^ 

q H 






Fig. 36. Joint F. 
k Ij d 

q not less than (.6 p -f- .4 d) 

Hence H " " " i/(i.i/ + .4 </) (.i/ + .4^) 
B = 6 d + 4 H. 

Thickness of each butt-strap not less than ^ the thickness 
of plate. 

The element: of the joint is A B D C, as in joint C, except 
that the rivets are in double shear. Taking, as before, the 


strength in double shear to that in single in the ratio 7 : 4, we 
have : 

Strength of Plate = t (p d) T 

Strength of Rivets -**. rcd~ S 

For equal strength of plate and rivets : 


i + 3-39 

n i 3-39 

-^- or n = i + 3-39 -y 

Efficiency = 


In this joint d -- t usually varies from I to ij4, as explained 
for joint E. These operations may be expressed by a rule 
similar to that for joint A, the numbered sections referring to 
that rule as given above. 

Rule : 

(1) Same as for joint A. 

(2) Use 3.39 instead of .645. 

(3)> (4), (5), (6) Same as for joint A. 

(7) Take ^/\ times the strength of one rivet, as found for 
joint A. 

(8), (9) Same as for joint A. 

Example. To lay out a joint as in F with I inch plates , 
using i 3-16 inch rivets. 

Then i -i. X 3-39 - i = 

And 4.03 + i = 5.03. 
And 5.03 X I -|-= 5-97 inches = pitch. 
Take pitch = 6 inches. 

Then with strength of plate at 60,000, as before, we have : 
Strength of plate in joint = (6 - - i-^-) X i X 60,000 = 

Area of i- 3 -inch rivet = 1.108. 


Strength of rivets = 1.108 X 5l X 60,000 X -821 =286,500. 
Original strength of plate 6 X i X 60,000 = 360,000. 
Efficiency = 286,500 -f- 360,000 = .796. 

Joint G. Double Butt-Straps. Rivets as in Joint D. 




; g 





Fig. 37. Joint G. 
A = Ij rf 

<7 not less than (.3 /> -f </) 

Hence;/" " " v (-S5/ + <0 (.<>5/ + <0 

5 = 6 rf + 4 //. 

Butt-straps to be of thickness not less than as given by the 
formula : 

Thickness of strap == g 5 , ~ , X (thickness of plate) 

5 (/ 2ft) 

The element of the joint is A B D C, as in joint Z), except 
that the rivets are in double shear. Taking, as before, the 
strength in double shear to that in single shear in the ratio 7 : 4, 
we have : 

Strength of Plate t (p d) T 
Strength of Rivets =^-nd 2 S 
. For equal strength of plate and rivets : 

A or 

+ 4-52 

Efficiency = 



In this joint d -f- t usually varies from about I to ij4, as 
explained for joint E. These operations may be expressed by 
a rule similar to that for joint A, the numbered sections re- 
ferring to that rule as given above. 

Rule : 

(1) Same as for joint A. 

(2) Use 4.52 instead of .645. 

(3), (4), (5), (6) Same as for joint A. 

(7) Take 7 times the strength of one rivet, as found for 
joint A. 

(8), (9) Same as for joint A. 

Example. To lay out a joint as in G with i l / 2 inch plates, 
using i^i inch rivets. 

Then ,| X 4.52 -H ii = I3 X */' X -'- 4-9 

X 3 

And 4.9 + i = 5.9. 

And 5.9 X i f = 9-6 inches = pitch. 

Take pitch for outer rows 9^3 and for inner rows 4-||- 

Then, with strength of plate at 60,000, we have : 

Strength of plate in joint = (9f - - i J) X ij X 60,000 = 


Area of iJ/ inch rivet = 2.074. 

Strength of rivets = 2.074 X 7 X 60,000 X -821 = 715,200. 
Original strength of plate = 9! X ij X 60,000 = 866,250. 
Efficiency = 715,200 -r- 866,250 = .826. 

Joint H. Double Butt-Straps. Triple Riveted, with 

double spacing in outer row on each side. 

h = ij d 
q not less than .3 p -f- .4 d 

g, " " " -3/ + ^ 

Hence H ^(.55 p + .4^) (.05 /> + .4 

Thickness of butt-straps found by same formula as for 
joint G. 

The element of the joint is A B DC, containing four whole 
rivets and two halves, or five in all. These are all in double 


shear. Taking, as before, the strength in double shear to that 
in single in the ratio 7 : 4, we have : 

Strength of Plate = t (p d) T 

Strength of Rivets = -2|_;r d 2 S 
For equal strength of plate and rivets : 

p d 

-^-or= i + 5.64 


n i 5.64 
n ~ a 4. 5.64 








Fig. 38. Joint H. 

In this joint d-- t usually varies from i to i%, as explained 
for joint E. These operations may be expressed by a rule sim- 
ilar to that for joint A, the numbered sections referring to that 
rule as given above. 
Rule : 

(1) Same as for joint A. 

(2) Use 5.64 instead of .645. 

(3). (4), (5) ( 6 ) Same as for J int A - 

(7) Take 8^4 times the strength of one rivet, as found for 
joint A. 

(8), (9) Same as for joint A. 

Example. To lay out a joint as in H with i^J inch plates, 
using I 7-16 inch rivets. 


And 5.9 + i = 6.9. 

And 6.9. + i -^-= 9.92 inches =. pitch. 

The limiting pitch by the. Board of Trade rule for this case 
would be 9.87. This means that a pitch of 9.92 or larger would 
not be passed without special permission. If necessary to re- 
duce below the limit, the joint should be re-designed with a 
smaller rivet. This case illustrates the point that these limiting 
values of the pitch, if rigidly adhered to, would prevent the at- 
tainment of the best joint efficiencies with thick plates. We 
shall here assume the right to proceed with the pitch derived 
from the formula which we will take as 10 inches for the outer 
and 5 inches for the inner rows. 

Then taking strength of plate at 60,000 we have : 

Strength of plate in joint = (10-- i-) X if X 60,000 = 

Area of i-- inch rivet = 1.623. 

Strength of rivets = 1.623 X 8f X 60,000 X -821 =699,600. 

Original strength of plate = 10 X if X 60,000 = 825,000. 

Efficiency = 699,600 -r- 825,000 = .848. 

Joint I. Double Butt-Straps. Triple Riveted, outer row 
on each side being double spaced, and passing through inside butt- 
strap only. 

q not less than .3 p + -4 d 

Hence H " " " T / (.55 p + .4 d ) (.05^ + .4 d) 

11 #i" " " i/(-55/ + ^)(-o5/ + ^) 
B=6d+ 2 H 

,= 64+ 2 H+ 2 H, 

Thickness of butt-straps found by same formula as for joint 
G. The element of this joint is A B D C, with four rivets in 
double shear and one in single shear. Taking, as before, the 
strength in double shear to that in single in the ratio 7 : 4, we 

Strength of Plate t (p d) T 
Strength of Rivets = 2 xd 2 S 
For equal strength of plate and rivets, we have : 



/ <: d 

-^- or;/=i -f 5.16 


" i 5.16 

a 4- 5- 



^ B, 




H q 

H, i "q, 




(.-- _.n__ --H, 



! ( 









i \ 














Fig. 39. Joint I. 

In this joint d -f- t usually varies from I to 1^4, as explained 
for joint E. These operations may be expressed by a rule sim- 
ilar to that for joint A, the numbered sections referring to that 
rule as given above. 

(1) Same as for joint A. 

(2) Use 5.16 instead of .645. 

(3)> (4), (5), (6) Same as for joint A. 

(7) Take 8 times the strength of one rivet, as found for 
joint A. 

(8), (9) Same as for joint A. 

Example. To lay out a joint as in I with i^ inch plates, 
using i 7-16 inch rivets. 

Then i-- x 5> l6 -^ 

16 ii 


And 5.40 + J = 6.40. 

And 6.40 X i -^- = 9- 2 inches = pitch. 


We will take 9^ for pitch of outer row, and hence 4^ for 
pitch of inner rows. Then, taking strength of plate at 60,000, 
we have : 

Strength of plate in joint = (9^ l^-) X if X 60,000 = 


Area of i-^- inch rivet 1.623. 

Strength of rivets = 1.623 X 8 X 60,000 X -821 = 639,600. 

Original strength of plate = 9J X if X 60,000 = 763,125. 

Efficiency = 639,600 -f- 763,125 = .838. 

An examination of the values of the efficiency will show 
that these various joints for the same value of d ~- t stand, in 
this respect, in the order : 

H, I, G, F, D, E, C, B, A. 


[i] Materials. 

Open hearth mild steel is used almost universally as the 
material for boiler construction, and in standard practice is used 
exclusively for shells, drums, heads, furnaces, combustion 
chambers and braces. Both steel and wrought iron are used 
for tubes, though solid drawn steel tubes may be considered as 
the better representing advanced engineering practice. 
Wrought iron is also used to some extent for rivets, though in 
the best modern practice steel rivets are preferred. 

[2] Joints. 

The various plates of a boiler are fastened together by 
riveted joints. These are of several varieties, as discussed in 
Sec. 15, and to which reference may be made. 

The holes in the plates are either drilled or punched. The 
former method is much the better. In the operation of punch- 
ing, a thin skin of metal about the hole is so severely strained 
that its strength, and especially its ductility and toughness, are 
reduced far below what they are in the remainder of the plate. 
This is not the case with the operation of drilling, or, at least, 
not to anything like the same extent. Drilled holes may also 
be located more accurately than punched holes, and thus with 
the former the parts of a riveted joint may be more perfectly 
fitted than with the latter. The operation of drilling leaves, 


however, a sharp edge, which should be removed by a reamer in 
order to avoid any tendency to cut the rivet. In spite of the 
greater cost of drilled holes they are now generally accepted as 
the best for all high-class work, and in many specifications no 
holes are allowed to be punched. 

Riveting is either by hand or by machine ; usually hy- 
draulic. The latter gives much the better result, and is pre- 
ferred where the machine can be made available. In many 
cases the construction is such that the jaws of the machine 
cannot be brought to bear on the joint, and in consequence hand 
riveting must be employed. 

After being riveted the joints are calked to insure tightness 
against leakage. This operation consists in beating down the 
edges of the metal against the face of the opposite plate by 
means of special pneumatic driven or hand tools, as shown in 
Fig. 40. These are known as calking tools, and are of two 




Fig. 40. Calking Tools. 

types, square and round nosed, as shown in the figure. The 
latter form is usually employed in modern practice. For op- 
erations on board ship the common hand tool is, of course, 
most commonly used ; but for extensive calking, as in the con- 
struction of boilers in the shop and where compressed air is 
available, the pneumatic driven tool is very largely displacing 
the hand tool. 

[3] Construction of Fire-Tube Boilers. 

We will now consider the chief features of the construction 
of a Scotch boiler. This will, at the same time, sufficiently il- 
lustrate the operations involved in the construction of other 
types of fire-tube boilers. 

In the best practice the longitudinal joints are double butt- 
strapped and triple-riveted in order to give to the boiler in this 
direction the highest possible proportion of the strength of the 
plate itself. The circumferential joints, those which run around 
the shell, are lapped and double or triple riveted. So far as in- 



ternal pressure is concerned the boiler is twice as strong to re- 
sist rupture around the girth as lengthwise so that a lapped cir- 
cumferential or girth joint is quite enough for strength alone, 
and it only remains to make it steam and water tight and to in- 
sure the necessary stiffness of the boiler as a whole. (See Sec. 
63.) Single-ended boilers are usually made with two courses 
of plates, as in Fig. n. Double-ended boilers are usually made 
with three courses. Each course consists of two or three 
sheets, varying with the diameter of the boiler. The heads are 
flanged, as shown in Fig. u, and thus secured by riveting to the 
shell. In some cases the shell has been flanged instead of the 
head, but such form of construction is rare. The head flanges 
are sometimes turned out, and sometimes in, as shown by the 
figure. Where machine riveting is to be used they must be 
turned out in order to allow the riveter to do its work. The 
back head is made usually in two pieces, with double or triple- 




Fig. 41. Styles of Corrugation. 

riveted lap joints. The front head is made in two or three 
pieces, according to size of boiler, usually with double-riveted 
lap joints. 

The furnaces, as shown, are corrugated in order to give 
greater strength and elasticity. There are three styles of cor- 
rugation in common use, as shown in Fig. 41. The furnaces are 
riveted to flanges formed on the front furnace sheet, and are 
connected by flanging to the sheets of the combustion chamber. 
Several different modes of connection are in use for this pur- 
pose. In one the furnace end is left plain and the flange is all 
on the combustion chamber sheet, as in Fig. 9. In another the 
combustion chamber sheets are left plain and the flange is on 
the furnace, as shown in Figs, u, 42, 44. In some forms pro- 
vision is made for removal and renewal without disturbing the 
furnace head sheets. Thus, in Fig. 9 the diameter at the front 


is the same as, or slightly larger, than that on the outside of 
the corrugations, and so the furnace may be withdrawn through 
the opening in the front sheet. In other forms of connection, 
where the furnace is flanged, especial provision must be made 
for removal, as shown in Fig. 43. 

Here the back end of the furnace is necked in on the bot- 
tom and sides, and a flange is thus obtained which only extends 
outside the outer diameter of the corrugation at the top. This 

Fig. 42. Flanged Furnace. 

flange serves to attach the furnace to the combustion chamber, 
and on cutting the joint loose the furnace may be taken straight 
to the front until the upper flange strikes the front sheet, and 
then swung upward and out of the front opening, as may be 
readily seen. 

In some cases where it is difficult to obtain the necessary 
room on the front head for the greater diameter of the outside 

Fig. 43. Removable Furnace. 

of the corrugation, or where, for other reasons, it is not con- 
sidered preferable to have the furnaces removable without dis- 
turbing the front sheet, the furnace end at the front runs out on 
the smaller diameter, as shown in Fig. 44. Some one of the 
forms favoring easy removal may be recommended as prefer- 
able in all ordinary cases. 

The combustion chamber, as shown, is built up of steel 
plates flanged and riveted together. The details of the con- 


stniction vary somewhat with the form of furnace attachment 
adopted, with the size of the boiler, and with the choice of the 
designer. The front plate is known as the back tube sheet. The 
top of the combustion chamber is sometimes flat, as in Figs. 9, 
65, and sometimes rounded up, as in Figs, n and 45. 

Fig. 44. Non-Removable Furnace. 

The tubes are secured into the tube sheets by "expanding," 
and "beading" or turning over at the back or at both the back 
and front ends. See Figs. 46, 47, 48. Tubes are expanded by 
means of a tool as shown in Fig. 49, representing the Dudgeon 
expander. The tool is introduced into the mouth of the tube 
and the small steel rolls are forced out by means of the tapering 
steel mandrel on which they rest. The mandrel is then turned 

Fig. 45. Rounded Top Combustion Chamber. 

around, and this by means of the frictional contact with the rolls 
causes them to turn also, and thus to roll around on the inner 
surface of the tube, carrying the whole tool slowly round and 
round. The mandrel is continually forced in and thus the rolls 
are forced otitward against the tube. The action is thus a roll- 


ing of the tube out against the tube sheet, and in this way the 
joint is made thoroughly tight. 

The Prosser expander, which was generally employed in 
former years, is now but rarely used. It consists, as shown in 

Fig. 47. Tube End. 

Fig. 46. Tube End. 

Fig 50, of a hollow tapering plug divided up into separate ele- 
ments or sections which are held together by a steel band. 
These are forced outward against the inner surface of the tube 
by driving a. taper mandrel into the hollow between the ele- 

Fig. 48. Tube Ends. 

ments. The action of the expander is thus to force the metal 
of the tube out against the edges of the sheet in a form of cir- 
cular ridge as shown in Fig. 46. 

Beading over the tube ends is usually done with a tool, as 

Fig. 49. Roller Tube Expander. 

shown in Fig. 51, and the result is as shown in Figs. 46-48. In 
some cases the tube sheet is recessed out for the beaded end of 
the tube, as shown in Fig. 47. The front ends of the tubes, as 
shown in Fig. 48, are usually swelled slightly larger than the rear 
ends to facilitate removal. The thickness of the metal of plain 


boiler tubes is usually from 8 to 12 wire gauge, or from about 
.17 to .10 in. 

In addition to the plain tubes fitted as before described, 

Fig. 50. Prosser Tube Expander. 

stay-tubes are also frequently fitted. These are of extra heavy 
metal, usually about % m - thickness, and specially fitted to the 

Fig. 51. Beading Tool. 

tube sheets by screw joints, as shown in Fig. 52. These tubes 
act as stays between the tube sheets. Further reference to this 

Fig. 52. Stay Tube with Ferrule. 

point will be found under the head of bracing. When stay 
tubes are fitted, it is customary to bead over only the back ends 
of the plain tubes, as in Fig. 48. Not infrequently, however, no 
stay tubes are fitted, and in such case the plain tubes must be 
beaded over on both ends in order that they may securely sup- 


port the tube sheets. Instead of the ordinary form of boiler 
tube, the Serve tube of cross-section, as shown in Fig. 53, is 
frequently fitted. The ribs of metal reach down into the column 
of hot gas moving through the tube and furnish additional sur- 
face to absorb the heat and help it through into the water. The 
surface on the fire side is thus much greater than the surface on 
the water side, while with the plain tube it is somewhat less. 

Fig. 55. New Admiralty Ferrule. 

Fig. 53. Serve Tube. 

Such tubes usually show an increased evaporation per square 
foot of surface measured on the water side, of from 15. to 20 per 
cent. Their increased weight, however, offsets in a measure 
this increase of evaporative efficiency per square foot of surface. 
Reference may also be made at this point to the use of re- 
tarders in boiler tubes. These are long twisted strips of thin 
sheet steel, as shown in Fig. 54. They are simply laid in the 
tubes and serve to give the gases more or less rotary motion 

Fig. 54. Retarder. 

and to assist in throwing them outward against the surface of 
the tube. With forced draft and high rates of combustion the 
use of retarders has been accompanied with a marked increase 
of economy. In some cases both Serve tubes and retarders 
have been fitted, but the special advantages of the combination 
may be called in question. 

As a measure of protection for the back ends of tubes un- 
der forced draft, cast iron ferrules are sometimes fitted. Fig. 



52 shows the so-called Admiralty ferrule in place in a stay tube. 
In Fig. 55 is shown an improved type of ferrule which by rea- 
son of the air space is believed to act still more efficiently to 
protect the tube sheet than the form shown in Fig. 52. 

Bracing. We must now consider the bracing needed to 
make the boiler perfectly secure and safe under the pressures 

Fig. 56. Adamson Ring. 

Fig. 57. Bowling Ring. 

to which the various parts will be subjected. The general prin- 
ciples to be kept in mind are as follows : (a) Cylindrical surfaces 
subjected to pressure on the concave side are not helped by 
bracing. They must be made sufficiently strong by giving to 
the material a suitable thickness, (b) Cylindrical surfaces sub- 
jected to pressure on the convex side may be stayed like a flat 

Fig. 58. Main Head Brace. 

surface, or they may be stiffened by ribs running around them 
in planes at right angles to the axis, (c) Flat surfaces will sup- 
port themselves if their area is sufficiently small in relation to 
their thickness and to the load per square inch, and it follows 
that large, flat surfaces must be sub-divided into parts of such 
size that they may thus become self-supporting. 


As an illustration of (b), furnaces were formerly strength- 
ened in this way, and the long favorite Adamson ring, as shown 
in Fig. 56, or the Bowling ring, as shown in Fig. 57, may be 
taken as good illustrations of this mode of adding support to 
cylindrical surfaces loaded .on the convex side. The present 
corrugated furnace, especially the Purves type, as shown in Fig. 
41, may be considered as a further illustration of the same prin- 
ciple. In modern marine boilers, aside from the furnaces, this 
mode of support is chiefly used to stiffen the bottom of single 
combustion chambers where screw stay bolts could not be read- 
ily fitted, and also in some cases the curved tops of combustion 
chambers. See Fig. 45. 

Coming next to flat surfaces as referred to under (c), the 
necessary sub-division is provided by the fitting of braces con- 
necting the part to be supported to some point where the sup- 
port can be provided, or by connecting together two surfaces 
urged by the steam pressure in opposite directions, as for ex- 
ample the two opposite heads of a boiler, as shown in Figs. 9, 
ii. Occasionally also flat surfaces are aided by attaching to 
them stiffening ribs of angle or tee bar, as on the front tube 
sheet, between the nests of tubes, or between the tubes and the 

Plates which are subjected to the direct action of the fire, 
as in the furnace and combustion chambers, are made relatively 
thin. This is done because a thin plate transmits heat better 
than a thick one, and is subjected to less severe internal stresses 
due to the difference in temperature of its two faces. The thin- 
ner the plate, however, the less the area which will be self sup- 
porting. Hence the braces for thin flat plates are relatively 
small and closely spaced, while those for thick plates are larger 
and spaced at greater intervals. 

The main head braces are secured as shown in Fig. 58. A 
washer is fitted on the outside to increase the supported area, 
and a nut is fitted both inside and outside so that the joint may 
readily be made tight, and that the brace may, if needed, act as 
a strut against pressure from without as well as a tie against 
pressure from within. In some cases a relatively thin plate is 
supported by a brace connecting it to a thicker or perhaps to a 
double plate, or to a place not requiring support itself, but which 
furnishes a convenient point for carrying the load. In such 
case the attachment to the thin plate is often made, as shown in 



Fig. 59, in order the better to sub-divide and distribute the sup- 
port. In double-ended boilers, certain parts of the head, as 
for example those between the furnaces, are supported by 
braces running obliquely back to the shell and attached as 
shown in Fig. 60. It often thus happens that braces must run 
at a slight obliquity in order to connect the parts to be sup- 
ported with convenient points of support. Other instances are 

Fig. 59. Forked End Brace. 

often found in the braces connecting parts of the back tube 
sheet below or between the tubes to the boiler head. In all 
such cases, wedge-shaped washers, as shown in Fig. 61, must be 
fitted under the nuts in order to get a good bearing between the 
nut and the shell. 

The braces connecting the relatively thin plates of the com- 
bustion chamber to the back head and shell of the boiler and to 



each other, are fitted by screwing them through into both plates, 
as shown in Fig. 62. The ends are sometimes riveted over and 
sometimes fitted with nuts. In some cases thev are left thread- 


o oj 

Fig. 60. Flange Foot Brace. 

ed the entire length, in others the threads are raised on the ends, 
as in the main head braces. The latter practice is much to be 
preferred. These braces are commonly known as "screw 

Fig. 61. Oblique Brace. 

stays," or "screw staybolts." This mode of fitting enables the 
screw stay bolt to act both as strut and as tie, or to resist pres- 
sure in both directions. In some cases the older form of 
"socket bolt," as illustrated in Fig. 63 is still fitted. In such 


case the head is riveted and the part of the bolt between the 
plates is provided with a hollow "socket." This acts as a strut 
and as a protection to the bolt proper. In modern approved 
practice screw stay bolts are either hollow or are drilled in at 

Fig. 62. Screw Stay Bolt. 

each end (see Fig. 64), to a point well beyond the inner face of 
the supported plate. The expansion and contraction of such 
parts of the boiler often have the effect of bending these bolts 

Fig. 63. Socket Bolt. 

back and forth, and they may thus in time become broken off, 
the break naturally occurring near the thicker of the two sheets 
where the bolt is held more rigidly. If this should occur, or 

Fig. 64. Improved Screw Stay. 

if the bolt should become badly corroded or pitted, especially 
near the plate, warning will be given of the fact by the escape 
of water or steam, and proper means must be taken for replac- 
ing the bolt. In this way timely warning may be given of a 



condition of affairs, which if allowed to go unnoted, might re- 
sult in a collapse of the plate, or in a disastrous explosion of 
the boiler as a whole. 

The usual spacing of stays, such as that shown in Fig. 58, 

o o o o 


O O O O 

Fig. 65. Girder Brace or Crown Bar. 

and supporting plates not directly exposed to the hot flames or 
gases is from 14 to 16 inches between centers, while for screw 

Fig. 66. Flanged Manhole and Fitting. 

stays supporting plates more or less directly exposed to the fire, 
the spacing is usually between 6 to 8 inches. See also on this 
point Sec. 62. 

For the support of the top of the combustion chamber, gird- 



ers or crown-bars are used, see Fig. 65. The load is transferred 
by means of the bolts from the combustion chamber plate to the 

Fig. 67. Reinforce Plate. 

girder, while the latter is supported by the edges of the vertical 
plates forming the front and back of the chamber. 

These girders are made of two pieces of steel plate, usually 
from y 2 to 3.4 in. thick, bolted or riveted together with distance 


Fig. 68. Manhole and Fitting 
in Shell of Boiler. 

Fig. 69. Reinforce Plate. 

pieces between so that the bolts which take the load from the 
flat plate may pass up between them as shown in the figure. 

The combustion chamber is sometimes secured to the back 
head of the boiler, or in double-ended boilers the two combus- 
tion chambers are secured together by plate braces fitted as 
shown in Fig. 45. Such are usually called gusset braces. 

In the general internal arrangement of the tubes, furnaces 


and combustion chambers, care must be taken to allow, as far 
as possible, a ready examination of the various parts. In good 
modern practice a space of from 10 to 12 inches is left between 
the nests of tubes and between the tubes and shell, so as to al- 
low the passage of a man from the steam space down through 
these spaces to the furnaces. 

Manholes and Covers. For the purpose of entering, ex- 
amining and cleaning the interior of a boiler, man or hand holes 
are cut in the head or shell. These are then covered by man- 
hole covers, plates or doors, as they are variously called. These 

Fig. 70. Manhole Plate and Fittings. 

are secured by bolts and dogs, as shown in Figs. 66-70. The 
usual size of a manhole is n by 15 inches, which are the dimen- 
sions required by the U. S. rules. It is of oval or elliptical 
shape, so that the cover with its lip extending over the edge 
may be gotten in and out without difficulty. The joint is made 
on the inside in order that the pressure may tend to keep the 
joint tight. A hand-hole is entirely similar in shape and fitting, 
and is simply smaller in size. In order to provide local strength 
and stiffness and to help support the load which comes on the 
feet of the dog, and also when the hole is cut in the shell 


to restore in some measure the metal taken out, a 
reinforce ring of metal is fitted about the hole. Such 
a ring of cast steel for a hole in the shell is shown 
in Fig. 68. The inner face is planed, so that the joint with 
the cover is readily made. In order that the removal of the 
metal may affect as little as possible the strength of the shell, 
the longer axis of the hole should run around the boiler rather 
than lengthwise. For holes in the head of a boiler the metal is 
often flanged inward, as shown in Fig. 66, the joint being made 
against the dressed edge of the ring. Where a manhole or 
handhole comes close to through braces, as, for example, near 

Fig. 71. Furnace Front. 

the furnaces, the reinforce plate may be formed, as shown in 
Figs. 67 and 69. At the angles or corners the plate is of suffi- 
cient width to let the threaded end of the brace come through, 
and the outside nut is then jammed down on the ring as shown. 
For heavy pressure the fitting illustrated in Fig. 70 may be rec- 
ommended. The reinforce ring is of flanged steel, and the 
cover of steel plate also, somewhat thicker than the metal of the 
shell. An angle iron, as shown, is riveted to the cover, making 
a neat fit within the reinforce ring, and keeping the plate ac- 
curately to its seat. 

Furnace Fronts and Doors. The furnace front is a fitting at- 
tached to the mouth of the furnace, and carrying the furnace 



door. In Figs. 71, 72 a common form of arrangement is shown. 
The front consists of a steel plate forming the outer part, 
and made with lugs or a flange for attachment to the furnace. 
The opening for the door is formed, as shown, within this front 
or door frame as it is sometimes called. Attached to this frame 
and with a space between is a second plate of cast iron 
forming the inner wall. This is pierced with a large number of 
small holes, while the frame is provided with a smaller number 
of larger holes. These are provided for the purpose of admit- 
ting air to the furnace above the grate. The inner plate is sub- 
ject to the direct action of the fire, and although cooled some- 
what by the air passing through, it is liable to burn out from 

/vfv.f.n i\ , i 


Fig. 12. Detail of Furnace Front. 

time to time. It is for this reason that it is made as a separate 
piece, and so is readily replaced as occasion requires. The door 
is formed in much the same way as the frame, and is provided 
with holes in a similar fashion and for the same purpose. Often 
a small covered peep-hole is provided for examining the fire 
without opening the door. In some cases also a small opening 
is made through which a slice bar may be introduced for stirring 
or breaking up the fire without opening the door. A form of 
slide or gridiron is also sometimes fitted so as to control the 
amount of air entering above the grates. In some cases the 
doors and frames are made entirely of flanged steel plates in- 
stead of cast iron, while much variety exists in the arrangement 



of the holes for the introduction of the air. Certain special 
fittings necessary to adapt the 'furnace fronts and doors to the 
application of forced draft will be referred to at a later point. 

The ash-pit door usually consists simply of a plate of thin 
sheet steel provided with the necessary lugs and handles, and 
covering the front opening in the furnace below the grate bars. 
It is used chiefly as a damper in connection with closed stoke- 
hold forced draft. 

Grate and Bridge Walls. The general arrangement of the 
inside of the furnace is illustrated in Fig. 73. The grate 
extends from the front of the furnace to the bridge wall as 
shown. The bottom of the door frame extends back a little 

Fig. 73. Section of Furnace and Grate. 

way and drops down, forming a kind of shelf for the support 
of the front ends of the grate bars. In some cases this exten- 
sion of the door frame extends back some distance, forming the 
so-called dead plate, upon which bituminous coal may be piled 
when first fired, so as to provide for the gradual distillation and 
combustion of its gases. 

The great bars may be made in a large variety of forms. 
In Fig. 74 is shown the standard type of cast iron bar. There 
are usually two lengths of bar in the length of the furnace, sup- 
ported by the door frame in front and bridge wall at the rear, 
and by bearing bars in the middle. These latter in turn are sup- 


ported at their ends by attachment to the furnace. The bars 
are usually cast double, as shown, while for convenience in fit- 
ting grates of varying widths, a small number of single bars are 
usually provided. The width of air space between the bars is 
usually made about equal to the width of the bar, or about one- 
half of the entire grate area, although this proportion should 
vary somewhat according to the fuel in use. The surface 
of the grate usually slopes slightly from front to rear, 
from i in 24 to i in 12, covering the usual range of angle. Cast 
iron grate bars often have a shallow groove running along the 
top. This fills with ashes and tends to prevent the clinkers ad- 
hering to the grate. 

In addition to the type of bar shown in Fig. 74, square 
wrought iron bars running the whole length of the furnace are 
sometimes used, and there is a large variety of patent and spe- 
cial kinds of shaking grate. The purpose in grates of this char- 

Fig. 74. Grate Bar. 

acter is to provide means for breaking up and working the fire 
without the need of opening the door. Many of them accom- 
plish this end to a considerable extent, but the greater simplicity 
and cheapness of the plain cast iron grate, as in Fig. 74, insures 
for the latter a wide use, and it is still the favorite in ordinary 

Turning now to the bridge wall, a common arrangement is 
shown in Fig. 73. A casting extends across the back of the 
furnace and is supported by attachment at the sides. This sup- 
ports the back ends of the grate bars, as already referred to, 
and also a wall of fire brick which forms the back limit of the 
grate, and over which the products of combustion pass on their 
way to the combustion chamber. 

Instead of fire-brick, the use of cast iron for bridges is be- 
coming frequent in modern advanced practice. Such bridges 
are of ribbed or channeled form, and in use they become suffi- 



ciently covered with ashes to form a protection against the heat 
of the fire. 

Front Connections and Funnel. After leaving the tubes at 
the front end the gases and smoke must be guided to the base 
of the funnel. This is done by the front connections or smoke 
boxes and uptakes, as shown in the diagrams. Fig. 75 shows 
the connection made between two single-ended boilers and one 
funnel, used in common by both. The boilers are placed front 
to front in an athwartship fire room. Fig. 76 shows the connec- 
tions between one double-ended boiler and the smoke pipe. 

Fig. 75. Front Connections, Uptakes and Funnel Base. 

These connections are formed of sheet metal riveted up in two 
or more thicknesses with an air space or non-conducting ma- 
terial between. The term front connection refers more especially 
to that part of the passage directly in front of the tubes. This 
is provided with doors swinging upward to allow examination, 
cleaning and repair of the tubes. A swinging damper is often 
placed in the uptakes for controlling the draft as may be de- 
sired, especially where two or more boilers are connected to one 
stack. The funnel or stack is also made of sheet metal riveted 
up, and in good practice in two thicknesses with a considerable 



air space between. This tends to prevent loss of heat by radia- 
tion, and thus the temperature of the gases is kept as high as 
possible while in the funnel, as is necessary for good draft. It 
may be remembered that for boiler economy the temperature of 
the waste gases at the front connection should be as low as 
possible, while for the sake of the draft all further loss of heat 
while in the funnel should be prevented. Around the base of 
the funnel is fitted an additional air screen or passage, 
known as the air casing. See Fig. 77. This serves to ventilate 





n n 

Fig. 76. Front Connections and Uptakes. 

the fire-rooms and to protect the neighboring parts of the ship 
from the heat radiated by the funnel. The air casing is pro- 
tected from the weather by a sloping ring of metal attached to 
the funnel, as shown in the figure, and known as the umbrella. 

The weight of the funnel is usually carried by straps or lugs 
attached to the structure of the ship, and it is furthermore stayed 
by guys on deck in order to provide the necessary steadiness 
and support in a sea-way. In small craft, however, the weight 
of the funnel is often taken simply by the uptakes and boilers. 



The funnel is often provided with a cover, which may be 
placed over the top when the ship is laid up, or when for other 
reason the funnel is not in use. The cover is usually kept a lit- 
tle distance above the top so as to allow the escape of smoke 
from small fires used for warming and airing the boilers. A 
ladderway should also be provided on the funnel to assist in ex- 
amination, adjustment of guys, fitting of cover, etc. In small 



Fig. 77. Funnel. 

craft a damper is often fitted in the funnel near the base, to as- 
sist in controlling the draft. 

[4] Construction of Water- Tube Boilers. 

Only a few points will require special notice under this 
heading. We have already seen that many types of water-tube 
boiler consist of one or more cylindrical drums above and one 
or more below, joined by a series of tubes. See Figs. 15-26. 
These drums, which are rarely more than 18 to 24 in. diameter, 

BOILERS. iog> 

are made from steel plates usually by flanging and riveting in the 
usual manner. The heads alone of such drums require con- 
sideration as regards bracing. If of sufficient size to require 
it they may be braced by through bolts as with boiler heads. 
In most cases, however, the heads are bumped or made of 
dished form, either concave or convex on the outside. The lat- 
ter is preferable, as the pressure is then carried on the concave 
side, and according to the United States law such heads are 
allowed without bracing a pressure the same as that for a 
cylindrical shell of a diameter equal to the radius of the sphere 
of which the head forms a part. 

It is much preferable to form the heads in this way, avoid- 
ing the need of bracing, and thus leaving the interior of the 
drum free for examination, at least so far as bracing is con- 
cerned. To allow access to the interior, manholes or hand- 
holes with appropriate cover plates are fitted to the heads. In- 
stead of forming these drums with riveted joints, drums with 
welded joints have recently come somewhat into use, especially 
in naval practice. 

With boilers having headers formed by the space between 
two parallel sheets, the necessary arrangements are quite differ- 
ent. These sheets require special support, and this is usually 
provided by screw stay-bolts or other equivalent stays worked 
between the two sheets attached to the tube sheet between the 
tubes as convenient, and securely tying the two sheets together. 
In some cases the outer sheets are supported by rod stays pass- 
ing from head to head through the tubes. In such case the 
tube sheets are left to be supported by the tubes which are thus 
thrown into compression, and the tubes must, therefore, be 
carefully expanded, especially on the inner side of the sheet, in 
order to give sufficient hold to support the sheets in this direc- 

In Sec. 14 [6] in speaking of the operation of water-tube 
boilers, reference was made to a separation in the upper drum 
of the water and the steam as they are delivered from the upper 
ends of the tubes. This is usually effected by some form of 
baffle plate, as illustrated in Fig. 17. A plate pierced with 
small holes is placed just in front of .the tube openings, and 
against this the escaping jets of water and steam are directed. 
The water is supposed to collect on the plate and to run down 
to the lower edge, or to the water in the lower part of the drum, 


while the steam passes through the holes and enters the steam 
pipe beyond. In the Bellville boiler, as in Fig. 26, there is a 
series of baffle plates forming a more or less tortuous passage 
through which the steam must pass on its way to the outlet, 
while at the last there is a plate with holes which exercise a 
straining action in the manner described just above. Special 
separators, as described in Sec. 35, are sometimes fitted in ad- 
dition to these internal separating devices. 

The tubes of water-tube boilers are of wrought iron or 
steel, and welded or solid drawn. For the bent-tube boilers 
solid drawn steel tubes are to be preferred. For straight tube 
boilers welded iron tubes are still in common use. The tubes 
are secured to the tube sheets either by expanding or by special 
fittings with screwed joints. In general there is a force tending 
to draw the tubes out of the tube sheet or junction box or 
other form of header, equal for each tube to its cross-sectional 
area multiplied by the steam pressure. This force must be re- 
sisted by the tube fastening, and while it is not usually serious 
in amount, its existence should not be forgotten, and the need 
of care in the fastening is shown. 

The furnaces of water-tube boilers are formed of grate-bars 
with a space below for ash-pit, all inclosed in the same general 
casing which surrounds the boiler as a whole, and as shown in 
the various figures referred to in the foregoing. 

Often a considerable amount of fire-brick is used as a lining 
to the furnace, and for protection to the lower ends of the tubes. 

Due to the great variety of forms of water-tube boilers, the 
details of construction often present the widest variation, and 
they cannot be so readily reduced to standard forms as in boil- 
ers of the fire-tube type. 

[5] Common Sizes and Dimensions of Scotch Boilers. 

The furnace diameter for Scotch boilers is usually found be- 
tween 42 and 48 inches. The upper limit comes about in the 
following manner. Taking into account the extreme ranges of 
temperature to which this part of the boiler is subjected, and 
based on general experience, it is usually considered that from 
Y-2 to y% inch is about as far as it is desirable to carry at present 
the thickness of the metal for the furnace. 

Again the strength of the furnace increases with the thick- 
ness and decreases with the diameter. Hence for a given pres- 
sure and limit on the thickness, the diameter will be limited as 


well. With modern pressures and a general limit on the thick- 
ness as above, the limit on the diameter, therefore, results. 
Rarely furnaces are met with up to 54 in., but only with the 
more moderate steam pressures. The lower limit for furnace 
diameter is given by a consideration of the necessary space be- 
tween the fire and the furnace crown. If this space or height 
is not sufficient the fires cannot be properly worked and the 
combustion will be incomplete, due to insufficient space for the 
admixture of air with the gases given off from the coal. 

In fact for efficiency of combustion we should probably pre- 
fer the diameter of the furnace larger than we are actually able 
to fit. As a lower limit, however, it may be considered inad- 
visable to fit furnaces much smaller than 42 inches, though they 
are sometimes found down to 36 inches. 

The length of fire grate is usually found between 5 and 6 
feet, though occasionally it extends to 6 ft. 6 in., or may be 
found as short as 4 ft. 6 in. The chief limitation here comes 
from the limit in the capacity of the average fireman to efficient- 
ly work his fire beyond a certain length. For average practice 
5 ft. 6 in. may be considered a good length, while it is doubtful 
if grate area added beyond this length will be of any great value 
for steam production. It is more than likely to become partially 
choked with ashes and clinker, while a shorter grate of 5 ft. or 
5 ft. 6 in. length may be kept bright and efficient over its entire 

The length of the furnace itself being equal to the tubes will 
be somewhat longer than the grate. The difference is usually 
from 12 to 24 or even 30 inches. 

This gyves for the usual length of furnace and of tubes from 
7 to 9 feet. 

The usual depth of the combustion chamber is from 24 to 
30 inches. This will usually give a suitable volume, and will 
also provide a sufficient space within which a man may swing 
a hammer or make use of such other tools as may be necessary 
in caring for the back ends of the tubes. 

The usual thickness of the water leg or space between the 
back of the combustion chamber and the back head of the boiler 
is from 6 to 9 inches. 

It will thus be seen that the usual length of a Scotch single- 
end boiler will be found between 10 and 12 feet ; 10 ft. 6 in. and 
IT ft. are quite common values. 


Comparing the construction of a single and double end 
boiler it is clear that the length of the latter will be slightly less 
than twice the length of a single end. This gives for the usual 
length of a double-end boiler from 18 to 21 feet. 

The diameter of the boiler will depend largely on the num- 
ber of furnaces to be fitted, and, of course, on the steam pres- 
sure and on any limitations in the thickness of the plates em- 
ployed. Modern four-furnace boilers are usually found between 
15 and 17 feet in diameter. For three-furnace boilers the di- 
ameters will similarly range from 13 to 14 feet, while for two- 
furnace boilers the diameter may vary from 10 feet or less to 
ii or 12 feet. 

The usual diameter for tubes is from 2% to 3 inches. The 
smaller sizes are used with forced draft and the higher rates of 
combustion. For natural draft 2^4 and 3 inches are common 

The thickness of boiler tubes is usually specified by sheet 
metal gauge number. Plain tubes are usually No. 8, 10 or 12, 
corresponding to .17 to .10 inch. Stay tubes are usually about 
No. 3 or J4 mcn m thickness. 

[6] Common Proportions for Scotch Boilers. 

Grate Surface (G. S.) 10 to 15 I. H. P. per sq. ft. G. S. 

Heating Surface (H. S.) 2 to 5 square feet per I. H. P. or 
25 to 40 square feet per sq. ft. G. S. 

Coal Burned. 15 to 30 Ibs. per sq. ft. G. S. per hour, or 
J / 2 to i lb. per sq. ft. H. S. per hour. 

Water Evaporated. 6 to 10 Ibs. per lb. of coal, or 4 to 10 
Ibs. per sq. ft. H. S. per hour. 

Section of Passage Over Bridge Wall. 1-6 to 1-8 G. S. 

Sectional Area of Tubes. 1-5 to 1-7 G. S. 

Sectional Area of Funnel. 1-6 to 1-8 G. S. 

Volume of Combustion Chamber. 3 to 4 cu. ft. per sq. ft. 
of G. S. 

Steam Volume. .3 to .4 cu. ft. per I. H. P. 

[7] Weights of Boilers. 

A modern four-furnace single-end Scotch boiler will weigh 
in the neighborhood of 40 tons, or upward, while the water will 
weigh not far from 20 tons, making 60 tons or more for 
the boiler as a whole. A four-furnace double-end boiler will 
similarly weigh not far from 70 tons, while the water will weigh 


not far from 40 tons, making no tons more or less for the boiler 
as a whole. For a modern three-furnace single-end boiler the 
weights would be similarly about 25, 15 and 40 tons, respec- 
tively, for boiler, water and total, while for a three-furnace 
double-end boiler they would be about 45, 25, and 70 tons, re- 
spectively, for boiler, water and total. Wide variations, of 
course, are found in the weights of boilers, and the above figures 
are only given to show the general nature of the weights in- 
volved. The weight of boilers with and without water, per 
square foot of heating surface, has already been noted in Sec. 
14. From these various figures and proportions it results that 
Scotch boilers may be expected to develop from 20 to 30 I. H. 
P. per ton according to conditions, while for water-tube boilers 
the figures will" run from 30 or 40 for the heavier types to 6c 
or 70 and even more for the lighter types, and with extreme 
rates of forced draft. 

[8] Western River Boat or Flue Boilers. 

As noted in Sec. 14 these boilers are in common use on the 
western rivers of the United States. A few additional points 
may here be given regarding their construction and installa- 

The length of such boilers varies from 20 to 30 feet, with 
a diameter of about 4 feet and with from 4 to 6 flues 10 to 14 
inches in diameter. The shells are made up of several courses 
as shown in Fig. 14, the circumferential seams being single riv- 
eted and the longitudinal seams double riveted. The flues are 
usually made also in lengths, lap welded and telescoped to- 

Such a boiler, for example, of 26 feet length and 47 inches 
diameter with six 10 inch flues has about 580 sq. ft. heating 
surface and will provide steam for about 275 I. H. P. in engines 
of the type commonly used and described in Sec. 24. 

When such boilers are arranged in battery they are placed 
side by side and are usually provided w^ith a single setting, thus 
giving a common furnace for the entire battery. The length 
of grate bar is short, being usually about 4 ft. 6 in. The boilers 
are furthermore usually connected by a steam drum on top and 
by one or more mud drums at the bottom. The steam drum 
for the size of boiler referred to above may be from 18 to 24 
inches in diameter, connected by legs 12 to 16 inches in diameter 


and spacing the boilers so as to give flame room of 9 to 12 
inches between the shells. 

In order to make the setting of such boilers as light as pos- 
sible the brick work may be kept down to a single thickness of 
fire brick supported by a sheet iron casing. The ash pan is 
preferably of steel plates lined with fire brick laid in cement. 
An interesting feature of the ash pan is the ash well which is 
frequently fitted. This consists of a 10 or 12 inch cylindrical 
passage leading from the surface of the ash pan down through 
the bottom- of the boat, and through which the ashes are dis- 
charged overboard without further handling. The boiler fronts 
are of cast iron with suitable fire-door and ash-pit openings. 
The uptakes, and funnels, or chimneys as they are more com- 
monly called, are made of heavy sheet iron and are supported 
by bracing carried down to the main deck beams which carry 
the boilers themselves. 

As noted below the engines work non-condensing, and the 
exhaust, as a rule, is led first through the feed heaters to the 
''Doctor" as described in Sec. 24, and then to the base of the 
chimneys for forming a blast and forcing the combustion. 
Connections may also be provided for carrying the exhaust to 
an exhaust pipe leading to the air, and also in part to the stern 
wheel if desired, in order to prevent the formation of ice in cold 

Lever safety valves, as illustrated in Fig. 78, are usually 
employed on these boilers, while gauge cocks, fusible plugs, 
steam gauges, and blow-off cocks are provided in accordance 
with usual practice. 

The steam piping is usually of lap-welded wrought iron with 
flanged joints. One of the chief features of western river prac- 
tice is the flexibility of the boat under different conditions of 
lading, and the necessity for allowing for such flexibility in the 
connections between the boilers and engines, and between the 
engines and wheel. Between the boilers and engines this is 
usually provided by the introduction of long bends or special 
connections intended to allow for changes due to expansion, 
contraction, twisting, etc. 

In addition to the flue boiler, as illustrated above, the di- 
rect fire tubular or locomotive type of boiler, as illustrated in 
Fig. 12, is sometimes used on the western rivers, each boiler be- 
ing thus self-contained, and the brick-work setting of the flue 


boilers being dispensed with. The flue boiler, however, is the 
more used and must be considered as the typical boiler in this 
field of practice. 


[i] Safety Valves. 

The purpose of the safety valve is to provide for the escape 
of the steam in case the pressure should tend to rise above the 
safe working limit for which the valve is set. There are two 
kinds of safety valves, known as lever and spring valves, accord- 
ing as the valve is kept down on its seat by a weight on a lever 
or by a powerful spring under compression. Fig. 78 shows the 
construction of the standard U. S. lever valve. 

The valve itself has a plain conical face, and fits to a corre- 
sponding seat as shown. In its motion up and down it is guided 
by the double stem, so that it can by no means become jammed 

Fig. 78. Standard Lever Safety Valve. 

in the chamber. The pressure of the steam comes on the bot- 
tom of the valve, and as it reaches or passes the limit for which 
the adjustment is made the valve lifts and the steam escapes 
about the edge, and thence is led by the escape pipe to the deck. 
The actual lift of a safety valve is very small, l /% in. being usually 
a large lift. The opening for the escape of the steam depends, 
therefore, on the circumference of the valve and on the lift, 
rather than on the area and lift. Safety valves are, however, 
usually designated and determined according to their area. The 
weight acts by means of the lever, as shown, and may be ad- 
justed so as to allow the valve to open at the pressure desired. 
On this point consult further Sec. 61. 

In modern practice the lever valve is infrequently used, ex- 
cept in vessels engaged in smooth water (river) service, the 
spring valve being fitted almost universally. In this form of 



valve, which is shown in Fig. 79, the chief point of difference is 
in the substitution of the spring for the weight and lever. The 
tension of the spring is adjusted by a screw at the top, so that 
the valve will not open until the limiting pressure is reached or 
exceeded. It is readily seen that as soon as the valve rises the 
spring is compressed and the tension is increased. It is also 
found that with the plain form of valve, as shown in Fig. 78, the 


Fig. 79. Double Spring Safety Valve. 

pressure on the face decreases the instant the valve lifts. Due 
to these facts it follows that such a valve, especially when con- 
trolled by a spring, is apt to seat itself the instant after rising, 
lifting again the next instant in answer to the restored value of 
the pressure. This irregular action will lead to a rapid opening 
and closing of the valve, producing a chattering noise very un- 
desirable in passenger boats, and interfering with the continuous 
and regular escape of the steam. To avoid this a lip of one 



form or another, as shown in Figs. 79, 80 and 81, is fitted to the 
valve at or beyond the edge, so that it may catch the escaping 
jet of steam, and thus increase the effective area of the valve 
after it has lifted from the seat. In such case the valve is forced 
farther from the seat, and while it still vibrates, it remains defi- 
nitely open until the pressure has fallen some 4 or 5 Ib. below 
that for which it opens. The valve then touches the seat in one 
of its vibrations downward, and remains closed until the pres- 

Fig. 80. Enlarged Section of Lip. 

Fig. 81. Safety Valve and Muffler. 

sure again rises to the point for which it is set. The safety 
valve s'hould always be fitted with a hand lifting gear, so that it 
may be opened by hand when desired, and the spring adjustment 
should be protected by lock and key, so that it cannot be 
changed by unauthorized persons. 

For large boilers, instead of one large valve, safety valves 
are often fitted in groups of two or three. This reduces to the 
smallest possible limit the danger from sticking or other de- 
rangement of the valve. The safety valve or valves should al- 
ways be attached to a fitting leading direct to the boiler, and 


with no possibility of closing it off by a stop valve. If both 
stop and safety valves are attached to the same fitting the'latter 
must always be placed inside or nearer the boiler than the 

[2] Muffler. 

This fitting, though not in the fire-room, may be properly 
referred to at this point. It consists of a metal chamber filled 
with bits of metal or stone, marbles, wire-gauze, small spiral 
springs, or with thin plates in layers pierced full of holes and ar- 
ranged in staggered fashion so as to provide a series of zig-zag 
passages for the steam. The steam from the safety valves and 
escape pipe makes its way to the air through this chamber, the 
purpose of the filling being to muffle or deaden the noise, which 
might otherwise seriously interfere with the giving of orders on 
deck. Fig. 81 shows a combined safety valve and muffler, the 
latter with plates as above described. Such an arrangement 
would be applicable for small or open craft having but one 
boiler, such as launches, small yachts, etc. 

[3] Stop Valve. 

Each boiler is connected through a separate boiler steam 
pipe to the main pipe. The entrance of steam to this pipe is 
controlled by the boiler stop valve, which thus provides for the 
regulation of the supply of steam to the engine, and for closing 
the boiler off entirely from the main steam pipe if necessary. 
The usual type of valve employed is shown in Fig. 82, and con- 
sists of a valve disc guided to its seat by wings, and raised or 
lowered by its connection with the screw spindle and handle as 
shown. As also shown in the figure the valve seat with wings 
and guide for the valve is very commonly a separate piece of gun 
metal or bronze, specially fitted for its strength and wearing 

Commonly in warship practice, and to some extent in mer- 
cantile practice, such valves are made self-closing in case of 
rupture of the boiler. In fundamental principle such a valve is 
a form of non-return valve, as illustrated by the check valve of 
Fig. 83. The screw stem does not open the valve, but limits 
simply the extent to which the valve can open. A second plain 
stem passing through the first then allows of the valve being 
pulled open by hand, even if there is no definite difference of 
pressure to force it open. In case of accident which reduces 



the pressure back of the valve so that the rush of steam is in 
the reverse direction to its usual flow, the valve will be closed by 
this rush and held securely on its seat by the excess of pressure 
on its outer face, thus shutting off the injured boiler, and retain- 
ing the others intact for use. If such an arrangement is not 
fitted and the valve cannot be closed by hand, or until it can be 
thus closed, an entire battery of boilers may be thrown out of 

Fig. 82. Boiler Main Stop Valve. 

use by the rupture of any one of them, all of the steam formed 
escaping through the one opening. Such form of valve should 
be placed with the spindle horizontal, so that its own weight 
may not enter as a direct factor in the movement of the valve 
toward and from its seat. 

In warships, where the pipes or boilers may be pierced by 
the fragments of exploding shell, such a safety provision may be 
of the utmost importance and value. 



[4] Dry-Pipe, or Internal Steam Pipe. 

This is a pipe of relatively thin metal placed within the 
boiler, extending lengthwise, and close to the top of the shell. 
At the inner end it is closed, and at the outer end connects with 
the pipe leading to the safety valve chamber, stop-valve and 
boiler steam pipe. Along the top of the pipe are cut a large 
number of narrow slits, through which the steam enters the 


Boiler Check Valve. 

pipe. This arrangement has the effect of drawing the steam 
from the highest part of the steam space, and of straining out 
some part of the entrained water. A small hole in the bottom 
of the pipe provides for draining off the water which may grad- 
ually collect. 

The uniform draft of steam from the whole length of the 
boiler tends also to prevent the priming which might be caused 
by drawing it all from one point. 



[5] Feed Check Valve and Internal Feed Pipe. 

The water from the feed pump conies to the boiler through 
the feed pipe, and then at the boiler passes through the feed- 
check. This is a screw-down, non-return valve, as shown in 
Fig. 83. The valve itself is entirely disconnected from the 
spindle, and the latter simply limits the height to which the valve 
can rise, while by screwing down sufficiently, the valve may be 
forced shut and held there. This construction is adopted so 
that should the feed pump stop working or between the strokes 
of the pump there may be no escape of water backward from 
the boiler into the pipe. Two such check valves are usually 

Fig. 84. Combined Check and Stop Valve. 

fitted to each boiler, one connecting with the main and the other 
with the auxiliary feed pumps. 

In modern practice a stop valve is usually fitted between 
the check valve and boiler, in order that, if necessary for ex- 
amination or repair, the check may be shut off from communi- 
cation with the boiler. Such combined stop and check valves 
are frequently fitted in a single casing, the stop valve, of course, 
being placed next the boiler (see Fig. 84.) 

After passing through the check-valve the water enters the 
internal feed pipe, by which it is led to the point or points of 
delivery. The end of the pipe is usually closed, and the water 
is delivered through a large number of small holes distributed 



along the pipe. The delivery is usually below the water level, 
and often between the nests of tubes where it meets with the 
rising currents of water heated by them. In some cases it is 
led to the bottom of the boiler, where it mixes with the rela- 
tively cool water there found; but this plan cannot be recom- 
mended, as it retards rather than assists circulation. In some 
cases also the water has been introduced as a spray into the 
steam space, but, while this plan has some advantages, it has 
not met with general favor. 

Fig. 85. Blow-off Cock. 

In water-tube boilers the feed water is usually fed into the 
upper drum, whence it joins the circulation in the boiler as 
noted in the description of boilers of this type. 

[6] Surface and Bottom Blows. 

Cocks or valves and connecting pipes are fitted for blow- 
ing grease and scum, or mud sediment and water, out of the 
boiler into the sea Fig. 85. The surface blow consists of a 
valve or cock attached to an internal pipe lying just below the 
normal water level, and either perforated with holes or leading 
to a shallow open pan. Outside the boiler there is a discharge 
pipe leading to an outboard valve, through which the discharge 
is effected. The scum and grease which collect on the surface of 


the water may by means of this arrangement be blown out of the 
boiler, and thus disposed of. In early engineering practice the 
bottom blow was of great importance, as it was used not only to 
discharge mud and sediment, but also the relatively dense water 
in the boiler when blowing down to reduce concentration, or 
when emptying the boiler of water for purposes of examination 
or cleaning. In modern practice with the surface condenser 
and the evaporator, blowing off to reduce concentration is no 
longer necessary, and blowing the water out of a boiler with 
its own steam is no longer considered good practice. The 
preferable plan is to allow the steam to condense and the water 
to cool down, and to then run it into the bilge or remove it by 
pump connections suitably arranged. Due to these facts, bot- 
tom blow valves have been sometimes omitted. There may 
still, however, be occasion to use such valves for the discharge 
of mud and sediment, and, therefore, they are still quite gen- 
erally fitted. In any event, there should be some valve and pipe 
connected with the lowest part of the boiler, and through which 
it can be emptied in one way or another. 

Both surface and bottom blows are usually fitted to water- 
tube boilers, especially to those types consisting of upper and 
lower drums with sets of connecting tubes. The surface blow 
is for scum and grease, while the bottom blow is essentially for 
mud and sediment, and is often attached to a special mud-drum 
provided to collect such substances. 

In many cases the inner end of the surface blow pipe ter- 
minates in a shallow pan located near the low water level in the 
boiler. The water within this pan will tend to remain more 
quiet than that outside, and the scum and impurities will thus 
collect therein, ready for removal by the use of the blowv The 
arrangement thus serves as a collecting pan for the surface 
blow, and by most engineers is believed to be quite efficient for 
the purpose in view. 

In other cases the pipe terminates in a closed end and is 
provided with a number of longitudinal slits through which the 
scum is drawn from the surface of the water. 

The cross-sectional area of the bottom blow may be so 
proportioned as to give about one square inch for every 5 tons 
of water contained by the boiler, with perhaps somewhat larger 
area in the case of small boilers. The area for the surface blow 
may be usually made from y 2 to 1-3 that of the bottom blow. 



[7] Steam Gauges. 

The steam pressure within the boiler, or rather the excess 
of the pressure within over the atmospheric pressure without, 
is shown by some form of steam gauge, of which the best-known 
and most used are those employing a Bourdon tube. In Fig. 86 
is shown such a gauge and tube, the cross section of the latter 

Fig. 86. Bourdon Steam Gauge. 

being an ellipse as shown. When the inside of the tube is sub- 
jected to the pressure of the steam it tends to become round in 
section, and as a result of this the tube as a whole tends to 
straighten out. This carries the free end outward, and this 
movement, by means of suitable connections, is made to give 
motion to the needle. These gauges are graduated by compari- 
son with a mercury column or other form of gauge tester, or 
with a standard gauge which has been thus graduated. Steam 
should not be allowed to enter these gauges, as the change in 
temperature may affect the accuracy of the reading. To pre- 



vent this the pipe leading to the gauge is always provided with 
a loop or U bend, called a "goose neck," which serves as a trap 
for the water condensed beyond this point. In this way the 
Bourdon tube and part of the connecting pipe are kept filled 
with water, which in turn is acted on by the steam, and thus the 
pressure is indicated without the actual presence of steam within 
the gauge. Steam gauges require comparison with a standard 
gauge from time to time, in order to make sure that their indi- 
cations are correct. They are often provided in duplicate, and 

Fig. 87. Water Gauges. 

frequently one gauge, at least, is provided of sufficient range to 
allow of use in the hydrostatic boiler test. 

[8] Water Gauge and Cocks. 

The level of the water within the boiler is shown by a ver- 
tical glass tube connected to fittings at each end, which in turn 
connect the one with the steam space and the other with the 
water space. As shown in Fig. 870, the entire arrangement of 
glass and fittings is attached to a hollow mounting called the 
stand-pipe, water-column, or water-gauge mounting. To the top 
and bottom of this are attached pipes, one leading to the steam 
space at or near the top, and the other to the water space at or 
near the bottom. In conne.cting the pipe with the steam space 
care must be taken that the opening is not near a steam outlet, 
as the rush of steam past such an opening might disturb the 


pressure and render the indications inaccurate by showing a 
higher level of water in the glass than actually exists in the 
boiler. At the bottom of the mounting a drain cock and pipe 
are provided, so that the glass may be blown through and 
cleaned as occasion requires. Screw plugs are also fitted above 
and below in a line with the base of the tube, so that if necessary 
a wire and swab may be run through the glass. Instead of the 
connections, as shown in Fig. 870, and which are to be considered 
as preferable, the ends of the water column are sometimes con- 
nected by horizontal passages directly to the boiler, as in Fig. 
8/b, which shows the fitting attached to the curved shell of boiler. 
With such a mounting the level of water in the glass is more 
liable to fluctuation and disturbance due to rolling of the ship 
or to priming than with the arrangement of a. 

Gauge glasses are usually from 12 to 15 in. in length, and 
y% to J4 in. diameter. Due to the fluctuations in temperature 
and the accompanying expansion and contraction, they are 
liable to occasional breakage. To avoid danger or trouble from 
the escaping jet of water and steam, it is quite customary in 
modern practice to fit the connection carrying the ends of the 
glass with ball non-return valves, working on a similar principle 
with the safety stop valve described above. So long as the 
glass is in place and the pressure equalized, the balls by their 
weight remain away from the seat and leave the passages open. 
Upon the breakage of the glass, however, they are carried by 
the rush of water and steam, each against its seat, thus closing 
the openings and stopping the escaping jets of water and 

In addition to the gauge glass, small cocks, three or four in 
number, are usually provided. In some cases such cocks are 
attached to the mounting, and in other cases to the boiler itself. 
These cocks serve as a check on the gauge glass, or for use in 
case the glass is not to be depended upon. The glass is usually 
so adjusted that when the water is at the bottom it is still some 
3 or 4 in. above the level of the highest heating surface. The 
water cocks cover about the same vertical distance, though in 
some cases the lowest cock is placed nearly on a level with the 
top of the heating surface. On single-end boilers two such 
water gauges are often fitted, one on either side at the front, and 
with water cocks at the back. Similarly on double-end boilers 
three would be fitted, two on one end and one on the other. 



i 9 1 Hydrokineter. 

This is an appliance used to force the circulation of the 
water in the boiler, more especially when raising steam. It con- 
sists, as shown in Fig. 89, of a steam jet and series of nozzles 

Fig. 88. Hydrokineter. 

with frame perforated at the back for the entrance of water. 
The steam is furnished from another boiler, and by its inducing 
action a current of water is set up and driven along, as shown by 
the arrows. This arrangement is placed near the bottom of the 
boiler, and thus serves to drive out the cold water which tends 
to collect there, and which is only slowly heated by the opera- 
tion of natural circulation. 

[10] Hydrometer. 

The density of the water in the boiler is determined by an 
instrument known as the hydrometer, and shown in Fig. 90. It 




may be of either glass or metal, and consists essentially of two 
bulbs with stem as shown. The upper and larger bulb is filled 
with air, and serves to give buoyancy to the instrument; 
while the lower and smaller bulb is weighted and keeps it in the 


upright position. When a body floats freely, wholly or partly 
immersed in a liquid, the weight of the body equals the weight 
of the liquid displaced. Hence, in this case the denser the water 
the less the volume displaced, and the higher the stem out of 
water. Average sea water contains about i part in 32 of solid 
matter, and hydrometers are usually graduated relative to this 
as a unit. That is, 2 means twice as much solid matter relative- 
ly as sea water ; 3, three times as much, etc., while o, of course, 
means fresh water'. The density of water depends, furthermore, 
on its temperature, so that the scale on the hydrometer can only 
be used with the temperature for which it was graduated. This 
is usually 200 deg. F., though frequently three scales are pro- 
vided; for 190 deg., 200 deg. and 210 deg., respectively. The 
water is drawn from the boiler through an appropriate pipe and 
connections into a deep, slender vessel called a salimometer pot. 
Soon after drawing, the water cools down through the tempera- 
ture corresponding to the hydrometer scales, and thus its den- 
sity is observed. 

[11] Boiler Saddles. 

The weight of the boiler is supported on saddles, or bearers, 
which in turn are attached to the structure of the ship. A 
modern form of boiler saddle is shown in Fig. 90, and consists 

Fig. 90. Boiler Saddle. 

of two or more supports on each side of the boiler of the lorm 
shown, and extending each one for some little distance longi- 
tudinally. An older form shown in Fig. 91 consists of a plate 
on edge connected with the structure of the ship, extending 
transversely under the boiler, and cut out to fit the round of the 



shell. The upper edge of this plate is fitted with angle irons on 
one or both sides to give a broader surface of support for the 
boiler. The form of saddle shown in Fig. 90 gives a better 
longitudinal support, and, moreover, makes access and exam- 
ination of the bottom of the boiler more easy than with the other 


Fig. 91. Boiler Saddle. 

form. Single-end boilers usually have two such saddles on each 
side, while double-end boilers are given three or four. 

In addition, the boiler is held in place in its saddles by stays 
adjustable by screw turnbuckles, or by other like means. A 
knee-piece, or chock, is also often riveted to the structure of the 
ship, projecting just above the end of the boiler at the bottom, 
and thus preventing endwise motion. 

[la] Boiler Jigging. 

To prevent loss of heat by radiation the boiler is covered 
with non-conducting, non-combustible felting, which in turn is 
held on by iron straps, or in some cases by a complete covering 
of sheet metal. This covering is known as boiler lagging. 

Sec. 18. DRAFT. 

Draft is due to a difference in pressure between the uptakes 
or base of the stack, and the ashpits. Due to this difference the 
air is driven up through the grate, thus supplying the amount re- 
quired for combustion, see Sec. u [2]. We must first inquire 
what it is that causes this difference of pressure. To make the 
case simple let AB Fig. 92, denote a grate with burning fuel, and 
ACDB the funnel. Then the pressure downward on the top of 
the fuel will be equal to the weight of a column of air and gas 
of cross section equal to AB, and extending up to the limits of 


the atmosphere. The pressure upward at the bottom of the 
grate will be the regular pressure of the external air, and this 
will equal the weight of a column of air of the same cross-sec- 

Fig. 92. Showing Principle of Natural 

ig- 93- Draft Gauge. 

tion GH, and extending also up to the limit of the atmosphere. 
The difference in the weight of these two columns is seen to lie 
in their lower ends, the bottom of one being composed of hot 
gas and the other of common air. The difference in pressure 
will, therefore, equal the difference in weight between the col- 
umn of hot gas CDBA, and that of air EFHG. Actually the 
column of liot gas will extend some distance above the top of 
the funnel before losing its heat or mingling with the air, so that 
the real height of the column is greater than the funnel. This 
difference is, however, usually neglected, and the difference in 
pressure is usually taken as the difference in weight between the 
column of hot gas extending from the grates to the top of the 
funnel, and a like column of external air. The pressure per 
square inch will, of course, be likewise equal to the difference 
between two similar columns of one square inch section. This 
shows the conditions for producing the so-called natural or fun- 
nel draft pressure. 

In order that the combustion may proceed, however, it is 


not enough to produce simply a column of hot gas. Care must 
also be taken to provide for a free and full inflow of the outside 
air to the grates in order that the amount necessary for combus- 
tion may be on hand as required. 

Draft pressures are usually measured by an instrument 
known as a draft gauge. As illustrated in Fig. 93, it consists 
of a bent U- tube partly filled with water as shown, and with a 
scale between the legs. In use the two legs are connected by 
appropriate means to the two places between which the differ- 
ence of pressure is desired, as for example, the ash-pit and fun- 
nel-base, or external air and funnel-base. In the latter case one 
leg is open to the air and the other is connected by a flexible 
pipe or other similar means to the uptake or funnel-base. With 
equal pressure in both places the water will stand at the same 
height in both legs, but with a difference of pressure it will rise 
in one leg and fall in the other, the movement being toward the 
lesser pressure. The difference in pressure is then measured by 
the weight of a column of water equal to the difference in height 
between the two legs. This is usually read in inches, and hence 
draft pressures are usually expressed in inches of water. With 
ordinary funnel draft the pressure is usually from 54 to ^ or ^4 
inch, with assisted or light forced draft from y 2 inch to I inch, 
with forced draft on large ships from I to 3 inches, while on fast 
yachts and torpedo boats the pressure may rise to 5 or 6 inches 
or more. In this connection it may be well to remember that 
an inch of water pressure is equal to a pressure of about 2-3 oz. 
per square inch. 

Since as above explained, natural draft is dependent on the 
difference in weight between the hot gas in the funnel and the 
outside air, it follows that the lighter, and, therefore, the hotter 
the gas the stronger the draft ; also the higher the funnel the 
greater the difference and the stronger the draft. A strong nat- 
ural draft with moderate height of funnel requires, therefore, 
a high temperature of escaping gases, and since these carry 
away heat to the outside air, this means a loss of heat and hence 
of economy. Strong natural draft requires, therefore, either a 
very high funnel, or a very high temperature of escaping gases 
with the resulting loss in economy. The usual temperature of 
the gases in the funnel base is from 600 cleg, to 800 deg. Fah. 
At a lower temperature the draft will be very poor, while with 
a higher temperature, the increase of draft will be obtained at 


the expense of economy. With natural draft the rate of com- 
bustion will usually range from 12 or 13 to 20 Ib. of coal per 
square foot of grate surface, dependent on the quality of the coal 
and other circumstances. 

From the preceding it is clear that with natural or funnel 
draft the power which can be obtained from a square foot of 
grate surface soon reaches a limit, and under present conditions 
this is usually found at from 10 to 12 I. H. P. In order to ob- 
tain more power per square foot of grate area, or in general 
more per pound of boiler, some form of assisted or forced draft 
is necessary. In all cases where very high speed is required as 
in torpedo boats, fast launches, yachts, etc., the application of 
forced draft is a necessity, as the boilers required to develop 
the power under natural draft would occupy far more weight 
and space than could be assigned them. With assisted or mod- 
erately forced draft the power per square foot of grate surface 
may be raised to from 15 to 18 I. H. P., and if properly installed, 
without sacrifice of economy. W r ith harder forcing the power 
may be raised to 25 or 30 I. H. P. per square foot of grate sur- 
face, or even more in extreme cases, but necessarily at the ex- 
pense of a loss in economy. 

The immediate object of all forced draft appliances is to in- 
crease the difference in pressure between the ash-pit and uptake 
over what it would be with the funnel alone, at the same time 
taking care to provide for the full supply of air to the grates as 
required by the rate of combustion desired. To this end there 
are four fairly distinct means as follows : 

(1) Closed fire room. 

(2) Closed ash-pit. 

(3) Exhaust fans in the uptakes, or between them and the 

(4) Steam jets in base of funnel. 

In the closed fire-room system the air is forced by means 
of blowers into the fire-room, which is closed air tight except 
for the outlet into the furnaces. The fire-room is hence under 
a pressure greater than the other parts of the ship, and to enter 
or leave it, an air lock is necessary, as illustrated in Fig. 94. 
Small air valves are provided by means of which the pressure in 
the lock may be equalized with that on either side, as may be 
desired, and this being done the door may be opened. To leave 
the fire-room, for example, both doors of the lock being closed, 



the pressure inside is equalized with the fire-room and the door 
being opened the person enters and closes it behind. The pres- 
sure in the lock is then equalized with that outside, the door is 
opened, and thus exit is effected. The chief advantages of this 
system lie in the fact that the boilers are left unchanged as for 

Fig. 94. Showing Principle of Air Lock. 

natural draft, and the shift from one system to the other is read- 
ily made. The necessary structural arrangements are also 
sometimes more readily effected than for the other systems, es- 
pecially in warship practice, and this reason may in some cases 
largely determine the choice. Its chief disadvantages lie in the 
difficulty of making the fire-rooms air tight, in the necessity of 
fitting air-locks as above described, and in the more severe 
strain placed on the fire-room force than with other systems. 

In the closed ash-pit system the air is forced by means of 
blowers into conduits leading directly to the ash-pits and fur- 
naces, which are closed air tight from the fire-room. The latter 
are, therefore, under a pressure greater than in the fire-room, 
and if the furnace, doors were opened with the draft on, the 
flames and gas would be driven out into the fire-room. To pre- 
vent this the draft must be shut off when the furnace or ash-pit 
doors are opened, and to avoid accidents a locking arrangement 
is often provided which prevents the door from being opened 
while the draft is on, or the draft from being turned on till the 
doors are closed. The Hotvdcn forced draft, which is representa- 
tive of this system, provides also for heating the air by means 
of the waste furnace gases before it enters the ash-pits and fur- 
naces. The general arrangement for the Howden draft is 
shown in Fig. 95. In the uptake is fitted a nest of vertical tubes 
through which the furnace gases pass on their way to the funnel. 
The air from the blower is delivered into a conduit which leads 




across the front of the boiler and within which these tubes are 
placed, as shown in the figure. The furnace gases pass, there- 
fore, through the tubes while the entering air passes about them 
on the outside, and thus absorbs a part of their heat which would 
otherwise escape through the funnel. The heated air then 
passes downward to a kind of special front over the ash-pits and 
furnaces. From these passages openings controlled by either 
sliding or hinged valves lead into the ash-pits and into the fur- 
naces above the grates. The supply of air to the fire may thus 
be regulated, and the relative amounts delivered above and be- 
low the grate may be adjusted as required for the best combus- 
tion. The ash-pit and furnace doors are, of course, air tight to 
the fire-room, and arrangements may be provided for insuring 
the closure of the valves before opening the doors as above ex- 

Induced draft is represented by the Ellis and Eaves system, 
as illustrated in Fig. 96. A large exhaust fan is so placed as to 
draw the gases along the uptakes and discharge them to the 
funnel, thus producing the draft by means of a defect of pres- 
sure in the uptakes, rather than by an increase in the ash-pit or 
fire-room. Considering the uptakes and funnel base as repre- 
senting the main passage, the fan is so placed as to draw from 
the former and deliver into the latter. Between the inlet and 
delivery is a slide or damper in the main passage, while the fan 
inlet may similarly be closed off. When the blower is in opera- 
tion the former of these is closed while the latter is open and 
the products of combustion are thus drawn from the uptake and 
delivered to the funnel-base on the other side of the main slide. 
On the other hand, if it is desired to run without the fan the 
inlet may be closed off and the main slide or damper opened, 
thus giving the usual arrangement for natural draft. 

The entering air which in this case is supplied either by nat- 
ural ventilation or by special blowers is heated before reaching 
the furnaces usually by being drawn around nests of tubes 
through which the gases pass on their way to the fan. These 
tubes are sometimes arranged vertically in front of the boiler as 
in the Howden system, and as shown in the figure, and some- 
times horizontally in the spandrels over or between the boilers. 
The air thus heated is then led through passages at the side of 
the front connections to the ash-pits, and to the spaces around 
the furnace frames. The furnace and ash-pit fronts are closed 




from the fire-room, and a certain proportion of the air is ad- 
mitted above and below the grates according to the needs of the 
combustion. This system has the advantage of leaving the fire- 
room open and of working the ash-pit under practically atmos- 
pheric pressure. Furthermore all leaks between the fire-room 
and the fire side of the boiler are inward, and thus the fire-room 
is kept free from escaping gas, or from flame and gas when the 
furnace doors are opened. Its chief disadvantage lies in the 
large size and weight of fan necessary to handle the gases, as 
compared with the smaller size needed for the air alone in the 
closed fire-room and closed ash-pit systems. 

The action of steam-jets in the base of the funnel is to pro- 
duce a defect of pressure in the uptake, thus giving a form of in- 
duced draft. Turning the exhaust of a non-condensing engine 
into the funnel produces the same result, and may also be con- 
sidered as a form of induced draft. The latter arrangement is 
sometimes met with in tugs and other small craft, while the 
steam jet is much used throughout the whole range of tugs, 
yachts, launches, tenders and all forms of small craft. One of the 
special advantages of the steam jet is its readiness for use as soon 
as a small head of steam is formed, and independent of any 
special auxiliary machinery. It is, however, a wasteful and ex- 
pensive mode of obtaining increased combustion, and is only 
to be recommended when simplicity and the saving of weight 
and space are of more importance than economy of steam. 

In this connection it must not, of course, be forgotten that 
the operation of the blowers used in the closed fire-room, closed 
ash-pit, or induced systems of draft requires also the consump- 
tion of steam and hence of coal, so that in no case can forced 
draft be obtained without paying for it in one form or other. 
General experience shows, however, that for a given increase in 
the rate of combustion, the use of a steam jet requires more 
steam than blowers, so that the latter are to be preferred except 
in such special cases as are referred to above. 

Reference may also be made to the practice of introducing 
jets of air under considerable pressure into the combustion 
chambers or furnaces of steam boilers. The result of such an 
arrangement is two fold : 

(1) It places the air at the immediate point where it may 
be of the most value in aiding to complete the combustion. 

(2) Its introduction under pressure insures its thorough 


mingling with the gases and the latter with each other, and thus 
still further aids in bringing about the conditions necessary for 
complete combustion. The introduction of air in this manner 
has met with considerable favor in many cases where it has been 
tried, especially in certain forms of water-tube boilers where the 
volume available for the combustion of the gases before they 
pass among the tubes is often limited or insufficient in amount. 


In the present section extracts are given from the United 
States Rules relating to the design of Marine boilers. To give 
here the Rules entire would require more space than is avail- 
able, but those of chief importance are given, though in some 
cases the wording is slightly changed to aid in condensation. 

Rivet Holes to be Drilled. 

All boilers built for marine purposes after July i, 1898, 
shall be required to have all the rivet holes "fairly drilled" in- 
stead of punched. 

Pressure Allowed on Cylindrical Shell Boilers. 

Rule. Multiply one-sixth (1-6) of the lowest tensile 
strength found stamped on any plate in the cylindrical shell by 
the thickness expressed in inches or parts of an inch of the 
thinnest plate in the same cylindrical shell, and divide by the 
radius or half diameter also expressed in inches and the quo- 
tient will be the pressure allowable per square inch of surface 
for single riveting, to which add 20 per cent for double riveting, 
when all the rivet holes in the shell of such boiler have been 
"fairly drilled" and no part of such hole has been punched. 

Test Pressure. 

The hydrostatic pressure applied must be in the proportion 
of 150 pounds to the square inch to 100 pounds to the square 
inch of the steam pressure allowed. 

Butt Straps. 

Where butt straps are used in the construction of marine 
boilers, the straps for single butt-strapping shall in no case be 
less than the thickness of the shell plates; and where double 
butt straps are used, the thickness of each shall in no case be 
less than five-eighths (^) the thickness of the shell plates. 


Stays and 'Flat Surfaces. 

In allowing the strain on a screw stay bolt, the diameter of 
the same shall be determined by the diameter at the bottom of 
the thread. 

No braces or stays hereafter employed in the construction 
of boilers shall be allowed a greater strain than six thousand 
(6,000) pounds* per square inch of section, and no solid or hol- 
low screw stay bolt shall be allowed to be used in the construc- 
tion of marine boilers in which salt water is used to generate 
steam unless said screw stay bolt is protected by a socket. But 
such screw etay bolts without socket may be used in staying the 
fire-boxes and furnaces of such boilers, and elsewhere when such 
screw stay bolts are drilled at each end with a hole not less than 
Y% inch diameter to a depth of at least ]/ 2 inch beyond inside sur- 
face of sheet, when fresh water is used for generating steam in 
said boilers (to take effect on and after July I, 1898, on all boil- 
ers contracted for or construction commenced on or after that 
date). Water used from a surface condenser shall be deemed 
fresh water. The flat surface at back connection or back end of 
boilers may be stayed by the use of a tube, the ends of which 
being expanded in holes in each sheet beaded and further se- 
cured by a bolt passing through the tube and secured by a nut. 
An allowance of steam shall be given from the outside diameter 
of pipe. For instance, if the pipe used be i l / 2 inches diameter 
outside, with a i l /\. inch bolt through it, the allowance will be 
the same as if a i l / 2 inch bolt were used in lieu of the pipe and 
bolt. And no brace or stay bolt used in a marine boiler will 
be allowed to be placed more than io l / 2 inches from center to 
center to brace flat surfaces on fire-boxes, furnaces, and back 
connections ; nor on these at a greater distance than will be 
determined by the following formulas. 

Flat surfaces on heads of boilers may be stiffened with 
doubling plate, tees, or angles. 

The working pressure allowed on flat surfaces fitted with 
screw stay bolts riveted over, screw stay bolts and nuts, or plain 
bolt with single nut and socket, or riveted head and socket, will 
be determined by the following rule : 

When plates 7-16 inch thick and under are used in the con- 
struction of marine boilers, using 112 as a constant, multiply 

* This limit is understood to refer only to the use of iron as a material. 


this by the square of the thickness of plate in sixteenths of an 
inch. Divide this product by the square of the pitch or dis- 
tance from center to center of stay bolt. 

Plates above 7-16 inch thick, the pressure will be deter- 
mined by the same rule, excepting the constant will be 120. v 

On other flat surfaces there may be used stay bolts with 
ends threaded, having nuts on same, both on the outside and 
inside of plates. The working pressure allowed would be as 
follows : 

A constant 140, multiplied by the square of the thickness 
of plate in sixteenths of an inch, this product divided by the pitch 
or distance of bolts from center to center, squared, gives work- 
ing pressure. 

Flat part of boiler-head plates when braced with bolts hav- 
ing double nuts and a washer at least one-half the thickness of 
head, where washers are riveted to the outside of the head, and 
of a size equal to % of the pitch of stay bolts, or where heads 
have a stiffening plate either on inside or outside covering the 
area braced, will equal the thickness of head and washers, the 
head and stiffening plate being riveted together * * * * 
shall be allowed a constant of 200, rivets to be spaced by thick- 
ness of washer on the stiffening plate. Boiler heads so rein- 
forced will be allowed a thickness to compute pressure allowed 
of 80 per cent of the combined thickness of head and washer, or 
head and stiffening plate. 

Plates fitted with double angle iron and riveted to plate 
with leaf at least two-thirds thickness of plate and depth at least 
one-fourth of the pitch would be allowed the same pressure as 
determined by formula for plate with washer riveted on. 

But no flat surface shall be unsupported at a greater dis- 
tance in any case than 16 inches, and such flat surfaces shall 
not be of less strength than the shell of the boiler, and able to 
resist the same strain and pressure to the square inch. 

Steel stay bolts of a diameter of ij4 inches and not ex- 
ceeding a diameter of 2^ inches at the bottom of the thread 
may be allowed a strain not exceeding 8,000 pounds per square 
inch of cross-section. Steel stay bolts exceeding a diameter 
of 2y 2 inches at bottom of thread may be allowed a strain not 
exceeding 9,000 pounds per square inch of cross-section, but no 
forged or welded steel stays will be allowed. 

Any steel stay brace of the Huston type, or similar thereto, 


prepared at one heat from a solid piece of plate without welds, 
intended for use in marine boilers, to be allowed a strain ex- 
ceeding 6,000 pounds per square inch of cross-section, shall be 
tested as hereinafter provided for steel bars intended to be used 
as stay bolts ; and any brace formed in this way, with an area 
of cross-section of 1.227 an d not exceeding an area of 5 inches, 
may be allowed a strain not exceeding 7,000 pounds per square 
inch of cross-section ; exceeding this area, may be allowed a 
strain not exceeding 8,000 pounds to the square inch. 

All steel bars intended for use as stay bolts to be allowed 
a strain exceeding 6,000 pounds per square inch of cross-section 
shall be tested by the inspectors, in lots not to exceed fifty bars, 
in the following manner : Inspectors shall promiscuously select 
one bar from each lot and bend one end of such bar cold to a 
curve, the inner radius of which is equal to one and one-half 
times the diameter of the test bar ; and should any such test bar 
break in the bending process the lot from which the bar was 
taken shall not be allowed to be worked into stay bolts for mar- 
ine boilers. 

Corrugated Furnace Flues. 

Corrugated furnace flues constructed with corrugations 8 
inches from center to center, the radius of outer corrugation 
being not more than one-half of the reverse or suspension curve, 
the plain parts of the ends not exceeding 9 inches in length, 
made of plates not less than five-sixteenths of an inch thick, 
when new, corrugated with practically true circles, shall be 
allowed a steam pressure in accordance with the following 
formula : 

Pressure in pounds 

Where T = thickness in inches. 

D = mean diameter, in inches. 

The strength of all corrugated flues, other than described 
in the preceding paragraph, when used for furnaces or steam 
chimneys (corrugation not less than i l / 2 inches deep, and not 
exceeding 8 inches from centers of corrugation) and provided 
that the plain parts at the ends do not exceed 9 inches in length 
and the plates are not less than five-sixteenths inch thick when 
new, corrugated, and practically true circles, to be calculated 
from the following formula : 


14000 X T = pressure. 

T = thickness, in inches. 

D = mean diameter, in inches. 

Ribbed Furnace Flues. 

The strength of ribbed flues, when used for furnaces or 
steam chimneys (rib projections not less than i^ inches deep), 
and not more than 9 inches from center to center of ribs, and 
provided that the plain parts at ends do not exceed 9 inches, 
and constructed of plates not less than seven-sixteenths inch 
thick, with practically true circles ; and 

The strength of corrugated flues, when used for furnaces 
or steam chimneys, corrugated by sections with flanged ends 
overlapping each other and riveted with ^4 inch rivets, 2 inch 
pitch, corrugated projection not less than 2y 2 inches from inside 
of flue to outside of lap, and not more than 18 inches between 
centers of corrugation, provided plain parts at ends do not ex- 
ceed 12 inches in length, constructed of plates not less than 7-16 
inch thick, with practically true circles ; and 

The strength of ribbed flues, when used for furnaces or 
steam chimneys, when made in sections of not less than 12 
inches in length, measuring from center to center of said pro- 
jections, and flanged to a depth not exceeding 2^ inches, and 
substantially riveted together with wrought-iron rings between 
such flanges, and such rings have a thickness of not less than 
double the thickness of the material in the flue, and a depth not 
less than 2^ inches, when straight ends do not exceed 12 inches 
in length, shall, in each of the above cases, be calculated from 
the following formula : 

C = 14000, a constant. 

T = thickness of flue in decimals of an inch. 

D = diameter of flue in inches. 

P = pressure of steam allowable. 

C x T 

Formula : P = ^. 

When plain horizontal flues are made in sections of less 
than 8 feet in length and flanged to a depth of not less than 
2 1/2 inches, and substantially riveted together with wrought-iron 
rings between such flanges, and such rings have a thickness of 
not less than half an inch and a width of not less than 2^ inches, 


or, in lieu thereof, angle-iron rings are employed, and such rings 
have a thickness of material of not less than double the thick- 
ness of the material in the flue and a depth of not less than 2 l / 2 
inches, and substantially riveted in position with wrought-iron 
thimbles between the inner surface of the ring and the outer sur- 
face of the flue, at a distance from the flue not to exceed 2 
inches, with rivets having a diameter of not less than one and 
one-half times the thickness of the material in the flue, and 
placed apart at a distance not to exceed 6 inches from center 
to center at the outer surface of the flue, the distance between 
the flanges, or the distance between such angle-iron rings, shall 
be taken as the length of the flue in determining the pressure 
allowable, which pressure shall be determined in accordance 
with the following formula : 

89600 X T 2 

P = 

L X D 

Where P = pressure of steam allowable in pounds. 
T = thickness of flue in decimals of an inch. 
L = length of section in feet. 
D = diameter of flue in inches. 

All vertical boiler furnaces constructed of wrought iron or 
steel plates, and having a diameter of over 42 inches or a height 
of over 40 inches, and crown sheets of flat-sided furnaces, if 
made with a radius of over 21 inches, and all cylindrical shells of 
back connections having a radius of over 21 inches, shall be 
stayed as provided ***** f or fl a t surfaces. But the 
cylindrical shell or bottom of back connections may be stiff- 
ened by angles or tees secured with rivets spaced no more than 
6 inches from center to center, the distance from center of 
rivets at edge to center of tee, or from center to center of tees, 
not to exceed 24 inches, tees and rivets to be of suitable section 
for the pressure and radius of surface braced. And the thick- 
ness of material required for the shells of such furnaces shall be 
determined by the distance between the centers of the stay bolts 
in the furnace and not in the shell of the boiler ; and the steam 
pressure allowable shall be determined by the distance from 
center of stay bolts in the furnace, and the diameter of such stay 
bolts at the bottom of the thread. Where steam chambers are 
formed in such vertical boilers at the upper end thereof by a 
sheet in form of a cone between the upper tube sheet and upper 


head of such boiler, the pressure allowed shall be determined by 
the diameter of such cone at the central point between the tube 
sheet and upper head of such boiler. 

Steam chimneys or superheaters formed of a flue, with an 
inclosing shell, shall be built as follows : 

The outer shell subject to internal pressure shall be con- 
structed under rules governing the shells of boilers, without al- 
lowance for any bracing to lining or flue. 

For lanings. 

The lining of flue subject to external pressure shall be con- 
structed as follows : 

Plates under 30 inches in diameter shall be at least 5-16 
inch thick. 

Thirty inches and under 45 inches diameter, plates shall be 
at least ^ inch thick. 

Forty-five inches and under 55 inches diameter, plates shall 
be at least 7-16 inch thick. 

Fifty-five inches and under 65 inches diameter, plates shall 
be at least J^ inch thick. 

Sixty-five inches and under 75 inches diameter, plates shall 
be at least 9-16 inch thick. 

Seventy-five inches and under 85 inches diameter, plates 
shall be at least y% inch thick. 

Eighty-five inches diameter a corresponding increase in 
thickness of plate of 1-16 inch for every 10 inches increase in 

The linings of flues shall be braced as follows : 

On or for all boilers using salt water, carrying a steam pres- 
sure of 60 pounds and under per square inch, the lining shall be 
braced with socket bolts, with heads, and with ends of bolts 
threaded for nuts, with plate washers not over 12 inches be- 
tween centers (or equivalent) on the inside of the lining; bolts 
to be at least I inch diameter. 

On or for all boilers using salt water, carrying a steam pres- 
sure over 60 pounds per square inch, the lining shall be braced 
with socket bolts, with heads, and with ends of said bolts thread- 
ed for nuts, with plate washers not over 10 inches between cen- 
ters (or equivalent) on the inside of lining; bolts to be at least 
ij^j inch diameter, the diameter of the bolts to be determined by 
the diameter at the bottom of the thread of said bolts. 

On or for all boilers using fresh water, the lining may be 


braced as described for boilers using salt water, or as hereafter 
described (or equivalent thereto), viz., with iron or steel angle 
rings, properly riveted to lining, and properly connected to 
outer shell by plate braces. These plate braces shall be of suffi- 
cient number and width to make space between plates not over 
20 inches on the lining; the angle rings shall be at least 2 l / 2 
inches by 2 l / 2 inches on lining, 5-16 inch and $ inch thick; 

3 inches by 3 inches on linings, 7-16 inch and y 2 inch thick ; 3^ 
inches by 3^ inches on linings, 9-16 inch and ^ inch thick, and 

4 inches by 4 inches on linings, 11-16 inch or more in thickness. 
Provided, //<wrrrr, That lining of steam chimney, between 24 
inches and 32 inches diameter and 9/3 inch thick, and lining be- 
tween 32 and 46 inches diameter, 11-16 inch thick, may be used 
in lengths not exceeding 8 feet, without bracing. 

The pressure of steam to be allowed on linings shall be 
determined by the following formula, viz : 
Constant, 89600. 
D = diameter in inches. 
T = thickness in decimals of an inch. 
L length in feet. 
P = pressure of steam allowable in pounds. 

89600 X T 2 

Formula : T ^ = P. 
i^ x D 

And the length of the lining or flue shall be the distance 
between center and center of angle rings, or center of angle 
rings to center of nearest row of rivets holding head, but in no 
case shall this distance be greater than 2.]/ 2 feet, except as other- 
wise provided. 

Corrugated or ribbed flues may be used as lining to steam 
chimney or superheaters under the same rules and conditions 
as apply to their use in the furnaces of steam boilers. 

Crown Bars. 
Working pressure = Ip X * D x L 

Where W = Width of combustion box in inches. 
P = Pitch of supporting bolts in inches. 
D = Distance between girders from center to cen- 

ter in inches. 
L = Length of girder in feet. 


d = Depth of girder in inches. 

T = Thickness of girder in inches. 

C 550 when the girder is fitted with one sup- 
porting bolt. 

C = 825 when the girder is fitted with two or three 

supporting bolts. 

C = 935 when the girder is fitted with four sup- 
porting bolts. 

Boiler Heads (Special Class). 

All heads employed in the construction of cylindrical boilers 
for steamers navigating the Red River of the North, and rivers 
whose waters flow into the Gulf of Mexico, shall have a thick- 
ness of material as follows : For boilers having a diameter ex- 
ceeding 32 inches and not exceeding 36 inches, not less than 
half an inch; for boilers exceeding 36 inches in diameter and 
not exceeding 40 inches in diameter, not less than nine-six- 
teenths of an inch ; for boilers exceeding 40 inches in diameter, 
not less than one-sixteenth of an inch additional thickness for 
every 8 inches additional diameter, required for boilers 40 
inches in diameter. 

And the heads of steam and mud drums of such boilers shall 
have a thickness of material not less than half an inch. 

Bumped heads may have a manhole opening flanged in- 
wardly, when such flange has sufficient depth and thickness to 
furnish as many cubic inches of material as was removed from 
the head to form such opening. 

Bumped Heads of Boilers. 

Multiply the thickness of the plate by one-sixth of the ten- 
sile strength, and divide by one-half of the radius to which head 
is bumped, which will give the pressure per square inch of steam 

To find the radius of sphere of which the bumped head 
forms a part, square the radius of the head. Divide this by the 
height of bump required. To this result add height of bump. 
This will give diameter of sphere, one-half of which will be the 
radius required. 

Unstayed Plat Heads. 

The pressure on unstayed flat-heads, when made of stamped 
material, on steam drums or shells of boilers, when flanged and 


made of wrought iron or steel or of cast steel, shall be deter- 
mined by the following rule : 

The thickness of plate in inches multiplied by one-sixth of 
its tensile strength in pounds, which product divided by the area 
of the head in square inches multiplied by .09 will give pressure 
per square inch allowed. The material used in the construction 
of flat-heads when tensile strength has not been officially de- 
termined shall be deemed to have a tensile strength of 45,000 

When such heads are stayed or braced, the pressure al- 
lowed shall be determined as above for flat surfaces. 

Pressure Allowable for Concaved Heads of Boilers. 

Multiply the pressure per square inch allowable for bumped 
heads attached to boilers or drums convexly, by the constant .6, 
and the product will give the pressure per square inch allowable 
in concaved heads. 


All manholes for the shell of boilers over 40 inches in di- 
ameter shall haye an opening not less than n by 15 inches in 
the clear, except that boilers 40 inches diameter of shell and 
under shall have an opening of not less than 9 by 15 inches in 
the clear in manholes. 

When holes exceeding 6 inches in diameter are cut in the 
boilers for pipe connections, man and hand hole plates, such 
holes should be reinforced, either on the inside or outside of 
boiler, with reinforcing plates, which shall be securely riveted 
to the boiler, * * * * * such reinforcing material to be 
of wrought iron or steel rings of sufficient width and thickness 
of material to equal the amount of material cut from such 
boilers, in flat surfaces ; and where such opening is made in the 
circumferential plates of such boilers, the reinforcing ring shall 
have a sectional area of at least one-half the area of material 
there would be in a line drawn across such opening parallel with 
the longitudinal seams of such portion of the boiler. On boilers 
carrying 75 pounds or less steam pressure a cast iron stop valve, 
properly flanged, may be used as a reinforce to such opening. 
When holes are cut in any flat surface of such boilers, and such 
holes are flanged inwardly to the depth of not less than i l /> 
inches, measuring from the outer surface, the reinforcement 
rings may be dispensed with. 


Also plates constructed of plate steel of corrugated form, 
without opening in plate for bolt, corrugation forming support 
for bolt, will be allowed for use for manhole and hand-hole 

No connection between shell of boiler and mud drum ex- 
ceeding 6 inches in diameter will be allowed. 

Safety Plugs. 

All steamers shall have inserted in their boilers plugs of 
Banca tin, at least one-half inch in diameter at the smallest end 
of the internal opening, in the following manner, to wit : Cyl- 
inder boilers with flues shall have one plug inserted in one flue 
of each boiler; and also one plug inserted in the shell of each 
boiler from the inside, immediately below the fire line, and not 
less than 4 feet from the forward end of the boiler. All fire-box 
boilers shall have one plug inserted in the crown of the back 
connection or in the highest fire service of the boiler. All up- 
right tubular boilers used for marine purposes shall have a 
fusible plug inserted in one of the tubes at a point at least 2 
inches below the lower gauge cock, and said plug may be placed 
in the upper head sheet when deemed advisable by the local in- 
spectors. All fusible plugs, unless otherwise provided, shall 
have an external diameter not less than that of a i inch gas or 
steam pipe screw tap, except when such plugs shall be used in 
the tubes of upright boilers, plugs may be used with an external 
diameter of not less than that of a three-eighths of an inch gas 
or steam pipe screw tap, said plugs to conform in construction 
with plugs now authorized to be used by this Board ; and it 
shall be the duty of the inspectors to see that these plugs are- 
filled with Banca tin at each annual inspection. 

Gauge Cocks. 

All steamers having one or two boilers shall have three 
suitable gauge cocks in each boiler. Those having three or 
more boilers in battery shall have three in each outside boiler 
and two in each remaining boiler in the battery ; and the middle 
gauge cocks in all boilers shall not be less than 4 inches above 
the top of the flues, tubes, or crown of the fire box. 

Safety Valves. 

Lever safety valves to be attached to marine boilers shall 
have an area of not less than I square inch to 2 square feet of 


the grate surface in the boiler, and the seats of all such safety 
valves shall have an angle of inclination of 45 degrees to the 
center line of their axis. 

The valves shall be so arranged that each boiler shall have 
one separate safety valve, unless the arrangement is such as to 
preclude the possibility of shutting off the communication of any 
boiler with the safety valve or valves employed. This arrange- 
ment shall also apply to lock-up safety valves when they are em- 

Any spring-loaded safety valves constructed so as to give 
an increased lift by the operation of steam, after being raised 
from their seats, or any spring-loaded safety valve constructed 
in any other manner so as to give an effective area equal to that 
of the aforementioned spring-loaded safety valve, may be used 
in lieu of the common lever-weighted valve on all boilers on 
steam vessels, and all such spring-loaded safety valves shall be 
required to have an area of not less than I square inch to 3 
square feet of grate surface of the boiler, except as hereinafter 
otherwise provided for water-tube or coil and sectional boilers, 
and each spring-loaded valve shall be supplied with a lever that 
will raise the valve from its seat a distance of not less than that 
equal to one-eighth the diameter of the valve opening, and the 
seats of all such safety valves shall have an angle of inclination 
to the center line of their axis of 45 degrees. All spring-loaded 
safety valves for water-tube or coil and sectional boilers re- 
quired to carry a steam pressure exceeding 175 pounds per 
square inch shall be required to have an area of not less than I 
square inch to 6 square feet of the grate surface of the boiler. 
Nothing herein shall be construed to prohibit the use of two 
safety valves on any water-tube or coil and sectional boiler, pro- 
vided the combined area of such valves is equal to that required 
by rule for one such valve. But in no case sha'll any spring- 
loaded safety valve be used in lieu of the lever-weighted safety 
valve without first having been approved by the Board of Su- 
pervising Inspectors. 

The first paragraph of this section applies to valves con- 
structed in material, workmanship, and principle according to 
the drawings for a safety valve printed with these rules, and all 
common lever safety valves to be hereafter applied to the boilers 
of steam vessels must be so constructed. 


Copper Steam Pipe. 

All copper steam pipes shall be flanged to a depth of not 
less than four times the thickness of the material in the pipes, 
and all such flanging shall be made to a radius not to exceed the 
thickness of the material in such pipes. And all such pipes shall 
have a thickness of material according to the working steam, 
pressure allowed, and such thickness of material shall be de- 
termined by the following rule : 

Rule. Multiply the working steam pressure in pounds per 
square inch allowed the boiler by the diameter of the pipe in 
inches, then divide the product by the constant whole number 
8000, and add .0625 to the quotient ; the sum will give the thick- 
ness of material required. 

The flanges of all copper steam pipes over 3 inches in di- 
ameter shall be made of bronze or brass composition, shall be 
securely brazed to pipe, and shall have a thickness of material 
of not less than four times the thickness of material in the pipes 
plus .25 of an inch; and all such flanges shall have a boss of 
sufficient thickness of material projecting from the back of the 
flange a distance sufficient to be properly riveted to the pipe, 
and of a thickness of not less than one-half inch ; and all such 
flanges shall be counterbored in the face to fit the flange of the 
pipe; and the joints of all copper steam pipes shall be made 
with a sufficient number of good and substantial bolts to make 
such joints at least equal in strength to all other parts of the 

Steel and Iron Pipe. 

The terminal and intermediate joints of all wrought iron 
and homogeneous steel feed and steam pipes over 3 inches in 
diameter, other than on pipe or coil boilers or steam generators, 
shall be made of wrought iron, homogeneous steel, or flanges 
of equivalent material ; and all such flanges shall have a depth 
through the bore of not less than that equal to one-half of the 
diameter of the pipe to which any such flange may be attached ; 
and such bores shall taper slightly outwardly toward the face of 
the flanges ; and the ends of such pipes shall be enlarged to fit 
the bore of the flanges, and they shall be substantially beaded 
into a recess in the face of each flange. 

But where such pipes are made of extra heavy lap-welded 
steam pipe up to and including 5 inches the flanges may be at- 


tached with screw threads ; and all joints in bends may be made 
with good and substantial malleable iron elbows or equivalent 

All feed and steam pipes not over 2 inches in diameter may 
be attached at their terminal and intermediate joints with screw 
threads by flanges, sleeves, elbows, or union couplings ; but 
where the ends of such pipes at their terminal joints are screwed 
into material in the boiler, drum, or other connection having a 
thickness of not less than y 2 inch, the flanges at such terminal 
joints may be dispensed with. Where any such pipes are not 
over i inch in diameter and any of the terminal ends are to be 
attached to material in the boiler or connection having a thick- 
ness of less than y 2 inch, a nipple shall be firmly screwed into 
the boiler or connection against a shoulder, and such pipe shall 
be screwed firmly into such nipple. And should inspectors 
deem it necessary for safety they may require a jam nut to be 
screwed onto the inner end of any such nipple. 

The word "terminal" shall be interpreted to mean the points 
where steam or feed pipes are attached to such appliances on 
boilers, generators, or engine, as are placed on such to receive 

All lap-welded iron or steel steam pipes over 5 inches in 
diameter or riveted wrought iron or steel steam pipes over 5 
inches in diameter, in addition to being expanded into tapered 
holes and substantially beaded into recess in face of flanges, 
shall be substantially and firmly riveted with good and sub- 
stantial rivets through the hubs of such flanges, and no such 
hubs shall project from such flanges less than 2 inches in any 

No cast iron nozzles, branch pipes, or elbows shall be used 
in connecting steam drums, superheaters, branch pipes, or 
steam pipes to boilers, and in no other part of steam pipes. 
Flanges welded to wrought iron, Bessemer, or other steel pipe 
may be used. No cast iron flanges will be allowed to be used 
on boilers for marine purposes unless such cast iron has been 
officially tested and test on record in the office of the local in- 
spectors where boiler with such appliances was constructed, 
and no cast iron with a tensile strength of less than 30,000 
pounds will be permitted to be used for such purposes. Semi- 
steel of not less than 24,000 pounds tensile strength may be used 
for nozzles, stop-valves, branch pipes, elbows, slip-joints, flanges 


to boilers, tee pipes, and water and gauge cock pipes or columns, 
when said semi-steel has been officially tested, and test on 
record in the office of the local inspectors, same as is required 
of cast iron. 

Coil and Ttibulous Boilers. 

All coil and pipe boilers hereafter made, when such boiler is 
completed and ready for inspection, must be subjected at in- 
spection to a hydrostatic pressure double that of the steam 
pressure allowed in the certificate of inspection to take effect 
on and after July I, 1897. 

The use of cast-steel manifolds, tees, return bends, or el- 
bows in the construction of pipe generators shall be allowed, 
and the pressure of steam shall not be restricted to less than 
one-half the hydrostatic pressure applied to pipe generators, 
unless a weakness should develop under such test as would ren- 
der it unsafe in the judgment of the inspector making such 

All drums attached to coil, pipe, sectional, or water-tube 
boilers not already in use or actually contracted for, to be built 
for use on a steam vessel, and its building commenced at or 
before the date of the approval of this rule, shall be required to 
have the heads of wrought iron or steel or cast steel, flanged 
and substantially riveted to the drums, or secured by bolts and 
nuts of equal strength with rivets, in all cases where the di- 
ameters of such drums exceed 6 inches. 

Except steam drums not exceeding 15 inches diameter at- 
tached to coils or pipe generators may be used when heads are 
made of malleable iron or cast steel, said drums being threaded 
on outside of such shell with a good full U. S. standard thread, 
eight to the inch, for a distance of at least I inch on such shell, 
the thread on head to correspond with the same and well fitted ; 
the end of shell projecting beyond the threaded part and screwed 
against a packing that will prevent water or steam to come in 
contact with the threaded part : 

PROVIDED, Such steam drums are placed outside of and not 
brought in contact with the heat or gases used in generating 
steam, and have been subjected to a hydrostatic pressure of 
double the steam pressure allowed. 

Drums and water cylinders constructed with bumped head 
at each or either end, any opening in the shell or heads to be re- 
inforced as required by the rules of the Board, the circumfer- 


ential and horizontal seams to be welded and properly annealed 
after such welding is completed, and when tested with a hy- 
drostatic pressure of at least double the amount of steam pres- 
sure allowed, may be used for marine purposes. 






The various types of marine steam engine may be classi- 
fied in different ways, according to the particular feature under 
special consideration. A typical modern marine engine as in 
Fig. 98 may be defined as a vertical, inverted, direct-acting, 
multiple-expansion, condensing, engine. Let us first examine the 
significance of these various terms. 

In the early days of marine engineering the engines were 
often horizontal, as shown in Figs. 101, 102, and such are still 
met with occasionally in special types of warship practice and 
elsewhere. An intermediate type, as shown in Fig. 103 and 
known as the inclined or diagonal engine, has been used to a 
considerable extent with paddle wheels. In modern practice, 
with rare exceptions, the marine screw engine is vertical, as in 
Figs. 97-100. 

In the earlier vertical marine engines the cylinder was at 
the bottom and the motion of the parts proceeded upward 
either directly to the crank shaft, as in the oscillating engine, 
Fig. 104, or to a beam or intermediate mechanism, Fig. 105,. 
whence it came back to the shaft. In the modern engine the 
cylinders are on top and the motion of the parts proceeds down- 
ward to the shaft. Hence in comparison with the earlier types 
the modern engine is called inverted. 

Where the connecting-rod and crank lie beyond the cross 
head or farther end of the piston-rod, as in Figs. 97-101, the 
engine is said to be direct-acting. In certain early types of hori- 
zontal engines in single screw ships, as represented in Fig. 102, 
the cylinder was sometimes placed close to the shaft and two 
piston-rods were fitted passing beyond the shaft, one above and 
the other below. Then from a crosshead at this point the mo- 



tion came back to the crank pin by a connecting-rod in the 
usual way. Such engines were called return connecting-rod or 
back-acting. In still earlier times the same type of engine placed 
on end, with the cylinder at the bottom, and known as the stee- 

Fig. 97. Longitudinal Section, Compound Engine, Mercantile Type. 

pie engine, was frequently fitted in side-wheel paddle steamers, 
and a modification of this is occasionally met with abroad at 
the present time. 

In early marine engines the expansion of steam always 





took place in one cylinder only. In the typical modern engine 
the steam is passed through a series of cylinders from one to 
another of increasing size. Such engines in general are termed 

Marine Engineering 

Fig. 99. Triple Expansion Engine, End View Looking Aft. 

multiple expansion. If the steam is thus used successively in 
two cylinders or the expansion occurs in two stages, the engine 
is said to be a compound; if in three cylinders or three stages, it 



is a triple or triple expansion; if in four cylinders or four stages, 
it is a quadruple or quadruple expansion, etc. 

Where the steam after being used in the cylinder is ex- 

Fig. 100. Triple Expansion Engine; End View Showing Condenser in 
Back Fr 


hausted into the air, the engine is said to be high-pressure or 
non-condensing. In the typical modern engine the steam is ex- 
hausted to a condenser, thus giving the advantage of an in- 




creased ratio of expansion and a decreased back pressure, 
engines are called condensing. 

Engines are often given special names according to the 
nature of the mechanical movements employed. In the usual 
type, as we have already seen, the motion is direct-acting and 
proceeds through piston, piston-rod, crosshead, connecting-rod, 

Fig. 101. Horizontal Direct Acting Engine, Outline. 

crank-pin and crank-shaft. In the beam engine, as shown in 
Fig. 105, the motion passes from the piston-rod to a crosshead 
and then by link or parallel motion to the beam. Thence from 
the other end of the beam it passes by the connecting-rod to 
the crank-pin and crank-shaft. Such engines are especially 
suited to side-wheel paddle steamers, and for many years were 

Fig. 102. Horizontal Back Acting Engine, Outline. 

considered the standard engine for use on river, bay and lake 
steamers. In more recent practice, however, the vertical direct- 
acting engine with screw propeller is to a considerable extent 
displacing the beam-engine with paddle-wheel, even in its own 

In the oscillating engine, a favorite in British practice for 
side-wheel paddle steamers, the cylinders are located below the 


shaft and are swung on trunnions, as shown in Fig. 104. The 
piston-rod is connected directly to the crank-pin, the piston-rod 
and connecting-rod forming thus but one member. This mo- 
tion is made possible by swinging the cylinder on trunnions, as 
may be readily seen by the diagram. The trunk type of hori- 
zontal engine, as shown in outline in Fig. 106, was often fitted 
in former years where economy of transverse or athwartship 
dimension was necessary. In this engine the use of the piston- 
rod was avoided by the large trunk, to which the connecting- 
rod was directly attached, as shown. 

The stern-wheel western river boat engine, as shown in 
Fig. 155, is a direct-acting horizontal engine connected to the 
stern wheel, and provided with a peculiar type of valve gear. 

Fig. ]03. Inclined Engine, Outline. 

Further reference to some peculiar features of this engine will 
be made in Section 22. 

The various members of a multiple-expansion marine en- 
gine may be arranged in a great variety of ways as regards the 
location of the cylinders, the crank angles, and the way in which 
the cranks follow each other around in the revolution. These 
are illustrated in Figs. 107-110. Of the many combinations 
which might be made, only the more important are men- 
tioned. Throughout these diagrams the high-pressure cylinder 
is denoted by //, the low-pressure cylinder by L, the interme- 
diate cylinder of a triple-expansion engine by /, and the first 
and second intermediates of a quadruple-expansion by/!,and/ 2 , 
respectively. Where the total cylinder volume is divided be- 



tween i\vo, each of half size, both of the latter are given the 
same letter. The course of the steam through the engine is also 
indicated by the arrows. For compound engines the usual ar- 
rangements are illustrated in Fig. 108. We may have two or 
three cylinders and one, two or three cranks. In the latter case 
the entire volume of low-pressure cylinder is divided between 
two cylinders, each of half the total volume. The first arrange- 
ment with high-pressure cylinder on top of low-pressure is 
known as a single-crank tandem compound, but is rarely met 

Fig. 104. Oscillating Engine, Outline. 

with in marine practice. The other arrangements may be 
placed, of course, with either end forward. The various crank 
angles are shown in Fig. 107 at I, 2, 3, 4 and 5, the crank 
marked / in No. 5 being in this case for one of the L. P. cylin- 
ders. With two cranks the angle between may be either 90 deg. 
or 180 deg., or slightly greater or less than 180 deg., as 175 
deg. or 185 deg. The 90 deg. angle is undoubtedly the best for 
all-around service. The 180 deg. angle gives a better balance 
to the moving parts and admits of a simplification of valve gear. 



and is sometimes preferred for these reasons. There is, how- 
ever, a liability of the engine's sticking on the center and the 
general readiness of handling is less than with cranks at 90 deg. 
To overcome this, angles of 175 deg. or 185 deg., as shown at 
2, are sometimes used, the balance of moving parts in such case 
being substantially as good as with an angle of 180 deg. 

Fig. 105. Beam Engine, Side Elevation. 

With three cranks the angles are usually equal, and hence 
120 deg. each. Occasionally they are slightly varied from these 
values in order to give a more uniform rotative effort, or to give 
a better balance to the forces causing vibration. 



For the triple-expansion engine the more important ar- 
rangements of cylinders are shown in Fig. 109. We may have 
three cylinders or more, and two, three or more cranks. The 

Fig. 106. Trunk Engine, Outline. 

most common types have either three or four cranks, in the lat- 
ter case the total L. P. volume being divided between two cylin- 


8 9 10 11 

Fig. 107. Various Crank Angles. 

ders, each of half the total volume. The crank angles are 
usually 1 20 deg. with three cranks, and 90 deg. with four, 

; * 


/ ^^ 















Fig. 108. Cylinder Arrangements for Compound Engine. 

though occasionally slight variations from these values are 
adopted in order to obtain a better balance of the forces causing 
vibration. Of the various arrangements of cylinders shown in 

i6 4 


Fig. 109, each may, of course, be placed either end forward in 
the ship. We may also have the various sequences and arrange- 
ments of cranks as indicated in Fig. 107, the changes of letter- 
ing where necessary being readily seen. 

For the quadruple-expansion engine the more important 
cylinder arrangements are shown in Fig. no. The number of 










x /h 

Fig. 109. Cylinder Arrangements for Compound Engine. 









Fig. 110. Cylinder Arrangements for Quadruple Expansion Engine. 

cylinders may be four, five or six, with four or five cranks. With 
five cranks the angles are usually equal, and hence of 72 deg.,. 
though as with three and four cranks slight departures might be 
made to obtain a better balance of the forces producing vibra- 
tion. The arrangements of cylinders shown in Fig. no may 
l>e placed in the ship either end forward, and various crank se- 


qucnccs in addition to those shown in Fig. 107 may be easily 
arranged. One of the chief tendencies of modern practice is to 
pay especial attention to the balancing of the forces producing 
vibration. The use of irregular crank angles in this connection 
lias been already referred to. In addition, and of not less im- 
portance, the larger cylinders with the larger and heavier pis- 
tons are now frequently placed inside, with the lighter moving 
parts on the outside, as in Fig. 109, Nos. 3 and 5, or Fig. 
no, No. 2. 


In describing the principal parts of the typical modern 
marine engine, we may take first the stationary, and then the 
moving parts. [x] Cylinders. 

As shown in Figs. 97-100, the cylinders are at the top of the 
engine and consist each of a cylindrical chamber containing the 
moving piston. The steam is received from the steam chest 
alternately in either end and thus forces the piston up and down. 
The motion is then transmitted through the piston-rod and con- 
necting-rod and thus the revolution of the crank and the crank- 
shaft is produced. 

Cylinders are made of cast iron of the highest grade. Each 
one, as shown in the figures, consists essentially of a cylindrical 
body or barrel, with which is usually cast the lower or bottom 
head. With the barrel are usually cast also the valve casings 
and chests and all ports and passages, as well as the necessary 
feet for attachment to the columns, lugs for attaching braces, 
etc. The top head or cover is cast separately and is secured to 
an appropriate flange on the barrel by means of stud-bolts. In 
some cases the head is made in a single thickness, conical in 
form to correspond to the piston, and ribbed on top for 
strength. In other cases it is made by a double shell or in two 
thicknesses with connecting ribs between. The lower head is 
formed in the same general way, but, as noted above, is usually 
cast in one piece with the barrel. 

In many cylinders, as shown in Fig. in, liners are fitted 
within the barrel or cylinder proper. These are of extra hard 
and fine grained iron, and are fitted for one or both of the fol- 
lowing purposes: (i) To provide a working surface admitting 
of replacement in case of excessive wear. (2) To provide a 



jacket space between the barrel and liner in case the cylinders 
are to have steam jackets. The space thus formed is filled with 
steam from the boiler, thus providing a jacket or layer of steam 
entirely surrounding the steam cylinder. Such an arrangement 
is known as a steam jacket, and is used to increase the economy 
of the engine as noted in Section 59. The liners are usually 
secured at the lower end by a flange, as shown in Figs, in, 113, 

Marine 'Engineering 

Fig. 111. Cylinder with Liner and Double Valve Chests. 

the joint between the end faces of the liner and barrel being 
carefully made in order to prevent leakage, especially if the 
space between the barrel and liner is to be used as a steam 
jacket. At the upper end the joint between liner and barrel 
may be made in a variety of ways. 

As shown in Fig. 112, a packing space is formed between 
the liner and barrel. This is filled with some form of elastic 



packing held in place by a ring attached to the liner as shown. 
In this way the tipper end of the liner is free to come and go as 
expansion and contraction may require, while the packing main- 
tains the joint steam tight. In another mode of fitting, a groove 
of dovetailed cross section is turned out partly in the liner and 
partly in the barrel, and a ring of soft metal or packing is ex- 
panded into the space thus formed. 

The bore of the cylinder or liner is made uniform except 
near the top and bottom, where it is counterbored out slightly 
larger, so that at the extreme ends of the stroke the piston 
rings may overrun the counterbore, and thus avoid wearing a 
shoulder in the metal. 

Fig; 112. Joint Between Liner 
and Barrel, Top. 

Fig. 113. Joint Between Liner and 
Barrel, Bottom. 

Cylinders as well as steam jackets are usually provided with 
drain cocks and valves with suitable piping, so that water collect- 
ing within them may be drained away In addition, automatic 
relief cocks or valves (see Sec. 24) should be fitted, set to open 
under an appropriate pressure, and thus furnishing relief in case 
a large quantity of water may find its way into the cylinder. 

The cylinders are supported directly upon the columns 
which are attached to facings on the lower head, or to lugs cast 
on the lower part of the barrel in case its diameter is not suffi- 
cient to reach out over the tops of the columns Sec Fig. 114. 
For mutual support the cylinders are quite commonly tied to- 



gather by braces, or flanged and bolted to each other. In some 
cases, however, the cylinders are allowed to stand alone and in- 
dependently, while in the other cases of recent practice a form 
of connection has been adopted, consisting of a vertical tongue 
and grooved joint. This allows differences of expansion ver- 
tically and fore and aft, but provides mutual support trans- 

Fig. 114. Double Inverted Y Columns. 

The valve chests with the various ports, passages, etc., are 
also cast with the cylinders, as shown in the figures. These 
parts will receive further notice in connection with valves. See 
Sec - 46. [a] columns. 

The columns serve to support the cylinders and to connect 
them with the bed-plate. They also serve to support the guide 
surfaces for the crossheads, and thus receive the transverse 



thrust of the connecting-rods. Columns are made either of 
cast iron, cast steel or forged steel. When of cast metal they 
are usually in the form of an inverted Y, as shown in Figs. 114, 
115, and of a box or I-formed section. When forged, the col- 
umns are usually cylindrical or slightly tapering, and sometimes 
hollow. Cast inverted Y columns both front and back of the en- 
gine, as shown in Fig. 114, for many years constituted standard 
practice. More recently, however, cast inverted Y columns at 
the back of the engine and cylindrical forged columns in front, 

Fig. 115. Inverted Y and Cylindrical Columns. 

as in Figs. 115, 116, are commonly employed in representa- 
tive marine practice. In such case either one or two columns 
may be fitted in front and one in the rear. When the columns 
are all cylindrical, it is customary to provide four for each cyl- 
inder. Such columns are usually placed vertical, as in Fig. 117, 
though occasionally they are spread somewhat at the base, as 
in Fig. 115. 

In some cases of modern practice four vertical columns of 
T section have been provided for use with a crosshead as shown 
in Fig. 134. The columns stand in pairs, one forward and one 



aft, and the wings of the crosshead carrying the slide surfaces 
work between them on the guides carried on their inner faces. 
In some cases the condenser is placed back of the engine 

Fig. 116. Inverted Y and Cylindrical Columns, Warship Type. 

and on the bed-plate, as in Fig. 100. In this case the back col- 
umns are either cast with the condenser shell or consist of short 
vertical columns standing on top of the condenser; which thus 
constitutes a part of the support of the cylinders as shown. 



To resist the racking and cross-breaking stresses to which 
the columns may be subject, it is necessary, especially with plain 
cylindrical columns, to provide transverse and even longitudinal 
ties and braces. The usual arrangement of such bracing is 
shown in Figs. 115, 116, 117. It will be noted in particular that 
the transverse bracing between a pair of columns as in Fig. 117 
unites them into a single girder, thus providing vastly more 
strength to resist lateral stresses due to rolling of the ship, etc., 
than could be furnished by the columns themselves and without 
the assistance which the bracing is able to provide. 

Fig. 117. Cylindrical Columns. 

The guide surface for a crosshead is fitted in various, ways 
according to the style of crosshead, the style of column and type 
of practice. The simplest arrangement is as shown in the cross- 
section of Fig. 118, in which the guide surface is fitted directly 
on the inner face of the Y column. In the arrangement of Fig. 
115 the guide surface is fitted on a separate slab of rather harder 
and finer grained cast iron, and hence better adapted for bear- 
ing purposes. Between this slab and the face of the column a 



space is left as shown, and through this may be circulated a 
stream of water to absorb the heat generated by the friction, 
and thus to keep the bearing surface cool. With cylindrical col- 
umns the guide c urface must be fitted as a separate slab for each 
crosshead, and usually in the manner shown in Fig. 117. These 
slabs may be of cast iron, steel or bronze, and are carried on 
longitudinal bars attached to the columns. The form of cross- 
section may be either hollow for water circulation, or plain or 
ribbed on the back for strength, as the case may require. A 
common form is that shown in Fig. 117, thinner towards the 

Fig. 118. Section of Cast Column Showing Guide Surface. 

ends and thicker in the middle as a girder, to provide the neces- 
sary strength at this point. 

For further details of the guide surfaces which depend on 
the form of crosshead used, reference may be made to [7]. 

[3] Bed-Plates. 

The purpose of the bed-plate is to support the feet of the 
columns, and thus to carry the weight of the cylinders and at- 
tachments, to provide seatings and support for the crank-shaft 
bearings, and generally to serve as the foundation piece upon 



which the engine rests, and through which its weight and the 
various stresses developed are transferred to the structure of 
the ship. 

As usually formed it consists of a series of transverse box 

Fig. 119. Bedplate for T 

Marine Enginttrlng 

pie Expansion Engine in one Casting 

or I girders, one for each crank-shaft bearing, these being con- 
nected together by fore and aft members, as shown in Figs. 
119, 121. 

Bed-plates are usually made of cast iron or cast steel. 

Fig. 120. Details of Bedplate in Fig. 119. Sections Showing Main Pillow Block. 

Rarely bronze or special forms of plate girder may be em- 
ployed. Large bed-plates instead of being made in one casting 
are often made in sections and bolted together. The bed-plate 
is secured to the ship by holding doivn bolts passing through the 



flanges of the plate and of the specially strengthened structure 
of the ship underneath, known as the engine seating or foun- 
dation. Further examples of bed-plates may also be noted in 
Figs. 97-100, 114, 116. 

o o o 


Fig 1 . 121. Bedplate in Sections. End View of one Section. 

[4] Engine Seating. 

This structure is a part of the ship, and serves to give the 
final support to the weight of the engine, and to lead the stresses 
due either to its weight or to its operation, into the structure of 
the ship as a whole. The usual character of the seating is 
shown in Fig. 122. It consists of a cellular construction formed 








Fig. 122. Engine Seating. 

by longitudinal and transverse vertical plates, stiffened and con- 
nected at the corners by angle irons, and usually forming a con- 
tinuous structure with a part, at least, of the regular internal 
members of the ship itself. 

We will now turn to the chief moving parts of the engine. 



[5] Pistons. 

The piston is the moving part of the engine upon which the 
steam directly acts, and which by the steam pressure is driven 
back and forth in the cylinder, and from which, through the 
piston-rod, crosshead and connecting-rod, the motion is trans- 

Fig. 123. Conical Marine Piston. 

mitted to the crank and crank-shaft. The requirements for the 
piston are therefore : (i) It must be able to support the load 
which the steam pressure brings upon it. (2) It must be of such 
form as to admit of movement up and down in the cylinder, at 
the same time making a steam tight joint between its outer edge 

Fig. 124. Marine Piston, Enlarged View, Showing Packing 
Rings and Follower Plate. 

and the cylinder walls. (3) Provision must be made for its se- 
cure attachment to the piston-rod, through which the forces are 
transmitted to the remaining moving parts of the engine. 

The usual form of marine piston is shown in Fig. 123, and 
consists of a shell of conical form with a central boss or body for 
carrying the piston-rod as shown. Around the outer edge of 


the piston the metal is thickened up to provide for the packing- 
rings, which are fitted to make a steam tight joint between the 
piston and cylinder walls. The fitting of these rings is shown in 
124. The rings are usually two in number, and are formed 



Fig. 1.25. Marine Piston, Joint in Packing Rings. 

of cast iron turned first to an outside diameter slightly larger 
than the bore of the cylinder.- They are then cut as shown in 
Fig. 125, and enough is taken out so that they may be sprung 
together sufficiently to allow their entrance into the cylinder 
bore. Care is taken to so locate the two rings that the cuts 

Marine Engineering 

Fig. 126. Marine Piston, Steel Springs for Packing Rings. 

shall not come opposite, and thus the opportunity for a direct 
leak through from one side to the other is avoided. In order to 
still further prevent such leakage, a tongue as shown in the 
figure is usually fitted across the opening. The tongue piece, 
which is usually of brass, is attached to the ring and overlaps 



the slit, as shown. The joints between the ring and tongue 
piece are carefully fitted so that in this way the ring may open 
and shut as circumstances may require, while the opening into 
the slit remains closed to the entrance of steam. When the pis- 
ton is of any considerable size it is customary to aid the natural 
elasticity of the rings by steel springs, as shown in Fig. 126. 
These bear on the bottom of the recess formed in the piston, 
and on the inner surface of the rings, and thus the latter are 
forced outward against the surface of the cylinder. 

The body of the piston itself, as shown in Fig. 124, is turned 
slightly smaller than the diameter of the cylinder, so that it 
clears the latter at all times, while the rings extend beyond and 
make the joint with the cylinder wall. The rings and springs 

Fig. 127. Ramsbottom Rings. 

are fitted as shown between the lower flange of the piston body 
an-1 a plate known as the follower plate or ring. By removing 
the latter the rings and springs may be removed when necessary 
for overhauling ana refitting. The follower plate is secured to 
the piston by stud bolts and nuts, as shown in the figure. In the 
best class of work all joints between the piston and rings, be- 
tween the follower plate and rings and between the two latter 
are carefully made by hand scraping and fitting, in order to re- 
duce the chances of leakage to the smallest possible limits. 

Many variations are met with in the details of the form and 
fittings of pistons. In some cases they are flat and either solid 
or hollow, as shown in Figs. 97, 128. 


In some cases ramsbottom rings are fitted instead of the 
rings of Fig. 123. These consist of two or three narrow rings 
turned slightly larger than the cylinder with a piece cut out so 
that they may be sprung on over the body of the piston, and into 
grooves, as shown in Fig. 127. No special springs are fitted, 
and the natural elasticity of the rings is depended upon to give 
the necessary pressure between the ring surface and the 

It is easily seen that no follower plates being fitted, the 
rings cannot be examined or removed without removing the 
piston. To avoid this difficulty the arrangement of Fig. 128 is 
sometimes used. Here the rings are carried on a larger solid ring, 
as shown, and sometimes known as a bull ring. This is carried 
between the faces of the piston flange and follower plate, and 
thus by the removal of the latter the whole arrangement may be 
withdrawn and examined. There is usually some clearance be- 
tween the inner surface of the bull ring and the body of the pis- 

Fig. 128. Piston with Ramsbottom Rings on Bull Ring. 

ton, as shown in the figure. This allows the whole arrange- 
ment of rings to move transversely independent of the piston 
body, thus making allowance for lack of alignment between the 
axis of the- piston-rod and the axis of the cylinder, or for wear 
in the latter. 

While light packing rings of cast iron fitted as above de- 
scribed without the assistance of steel springs may prove satis- 
factory for small pistons, the more standard method of Fig. 123 
is to be recommended for all cases where the pistons are of any 
considerable size. 

In present practice pistons of the form shown in Fig. 123 
are made of cast steel. Pistons of the form shown in Figs. 97, 
128, are more commonly made of cast iron. 

The chief advantage of the conical form of piston lies in the 
saving of weight for the necessary strength and stiffness, as 
compared with other forms. This superiority has gained for it 



almost universal adoption in modern practice, and it may be 
considered as the present day representative form of marine 
piston, and the one which will naturally be adopted unless there 
may exist special reasons for the adoption of the older type. 

[6] Piston-Rods. 

The piston-rod is that member of the moving parts which 
serves to support the piston, to carry the forces due to the 
steam pressure through the stuffing box outside the cylinder, 
and through the crosshead to communicate them to the connect- 
ing-rod and other moving parts. The requirements are there- 
fore as follows : (i) It must have sufficient strength and stiffness 
to safely carry the load coming from the piston. (2) It must 
be provided at the upper end for attachment to the piston, and 
at the lower end to the crosshead. (3) It must be of such form 
as to admit of readily making a steam tight joint where it passes 
out of the cylinder. To fulfil these conditions the piston-rod, as 







1 1 


!i;i;:. : !'!i 

: ' ! 

1 1 1 i ! ! i 

Marine Engineering 

Fig. 129. Piston Rod. 

shown in Fig. 129, has the form of a uniform cylindrical rod ex- 
cept at the ends where it joins the piston and crosshead. The 
common form of attachment to the piston is shown in the figure. 
The rod is sometimes tapered where it lies in the piston, and 
sometimes parallel. It is often relieved from direct bearing ex- 
cept near the top and bottom, so as to give definite points of 
bearing where it is most needed. A shoulder or ring is also 
fitted, as shown, so as to give a definite stop against which the 
Dody of the piston rests. The end of the rod is threaded and a 
nut on top completes the fastening. This nut is sometimes 
hexagonal and sometimes cylindrical with longitudinal grooves, 
a spanner wrench being used in the latter case to set the nut 

The fitting at the lower end of the piston-rod depends on 
the style of crosshead used, and may be more appropriately de- 
scribed under that heading. 

In modern practice with conditions requiring the highest 
grade of material and most careful design, the piston-rod is 



often made hollow. This practice also extends to most of the 
other cylindrical elements of the engine such as cylindrical col- 
umns, crosshead pins, connecting-rods, crank-pins, crank, line, 

Fig. 130. Marine Crosshead. 

thrust, and propeller shafts. Inasmuch as this style of construc- 
tion was first commonly introduced in connection with shafting, 
the reasons for such practice and its advantages may properly 
be discussed under that heading. 

Fig. 131. Marine Crosshead, Slipper Type with Cotter Fastening for 
Piston Rod. 

[7] Crossheads. 

There are several types of crosshead to be met with in 
marine practice. In Fig. 130 is shown one of the more com- 



mon forms. It consists essentially of a cubical body A, through 
which in a vertical direction is the hole for the piston-rod. Ex- 
tending out on either side longitudinally are the two crosshead 


Fig. 132. Marine Crosshead, Slipper Type. 

pins B and C. Then attached to the two remaining sides trans- 
versely are the slides as shown. The connection between the 
crosshead and the piston-rod is commonly by means of thread 
and nut, as shown in the figure, in the -same way as for the con- 

Marixt Engineering 

Fig. 133. Crosshead formed on Lower End of Piston Rod. 

nection to the piston. In some cases a pin or cotter joint, as 
shown in Fig. 131, is used instead of the thread and nut on the 
end. The slide surfaces D and E rest on the guide surfaces of 
the columns, as above described, one side taking the load when 



going ahead and the other when backing. A crosshead of this 
type is therefore suitable for double inverted Y columns where 
there is a guide surface on both back and front sides of the en- 
gine. Where cylindrical columns are used, as in Figs. 115-117, 
the slipper form of crosshead is commonly fitted. This is shown 
in Fig. 132, and so far as the part connected with the piston- 
rod and carrying the crosshead pins is concerned may be the 
same as in Fig. 130. Instead of two wings carrying slides, how- 
ever, there is but one with the form shown in the vertical view. 
The corresponding form of guide is shown in Figs. 115, 117. 

Marine Engineering 

Fig. 134. Marine Crosshead, Special Type. 

When going ahead the face A of the slipper bears against the 
face B of the guide. Cheek pieces or gibs C and D are se- 
cured to the column, thus forming guide surfaces on their in- 
ner faces E and F. Against these the faces G and H of the slip- 
per bear when in backing motion. The go-ahead surface is 
therefore formed on the faces A and B of the guide and slip- 
per, as with Fig. 130, while the backing surface, instead of be- 
ing provided on the opposite column and on another slide piece 
on the other side of the crosshead, is formed on the reverse 
side, G and H, of the slipper, and on the cheek pieces, as shown. 
These types of crosshead are suited to the so-called forked 
type of connecting-rod, as described below, for which the cross- 



head pins are a part of or fast in the crosshead, and are naturally 
two in number, one on either side fore and aft, as shown. In 
the other type of connecting-rod which is frequently met with, 
the rod is not forked, and the pin is fast in the upper end, and 
is single rather than double, while the crosshead member fur- 
nishes the bearing. This arrangement is shown in Fig. 133. 
The crosshead body is usually forged up on the lower end of the 
piston-rod, and together with a suitable cap and bearing brasses 
forms the bearing for the pin which is fast in the upper end of 
the connecting-rod, as shown. The wings for carrying the 
slides may be attached and the slides may be fitted in the same 


Fig. 135. Marine Crosshead, Special Type. 

general manner as in Figs. 130, 132, either double or of the 
slipper type. 

A third form of crosshead occasionally found in modern 
practice was referred to in [2], and is here shown in Fig. 134. 
This type of crosshead is a marine adaptation of a type very 
common in stationary engine practice. The slide surfaces are 
formed on the opposite faces of webs or wings extending out 
from the body of the crosshead, and bearing on the guide sur- 
faces formed on the columns. In Fig. 135 is shown a somewhat 
different form of the same type of crosshead, the latter being 
suited to two columns and the former to four. 

As noted in [6], the crosshead-pins, in the most advanced 
practice, are often made hollow. In some cases the hole is 

1 84 


parallel ; in others its diameter decreases from the outer end 
inward, thus giving the most metal at the inner end, where the 
greatest stresses are likely to be found. See Fig. 130. 




Marine Enyineeriny 

Fig. 136. Marine Connecting Rod. 

[8] Connecting-Rods. 

Fig. 136 illustrates perhaps the more common type of con- 
necting-rod. At the upper end it is forked or formed into a U 

Fig. 137. Marine Connecting Rod. 

shape, each branch being provided with a bearing and connec- 
tions for one of the crosshead-pins. This type of end corre- 
sponds therefore to the type of crosshead shown in Figs. 
130, 132. 



For connection to the crank-pin, the lower end of the rod is 
fitted with brasses and cap, all secured to the forged out foot 
of the rod by through bolts as shown. 

For the type of crosshead shown in Fig. 133 the rod is 
formed, as shown in Fig. 137, with a U-shaped upper end fitted 
to receive the two ends of the crosshead-pin, which is thus made 
fast to the rod. This pin is then seated in a bearing in the cress- 
head, as described under that heading. The lower end of the 
rod is usually of the same form as shown in Fig. 136. 

Rarely the gib and key form of connecting-rod end as illus- 
trated in Fig. 138, is found in marine practice. 

In external form marine conecting-rods usually increase 
in transverse dimension from top to bottom. In some cases 
they are given a uniform taper from one end to the other, as in 

Marine t-nginetring 

Fig. 138. Marine Connecting Rod with Gib and Key Connections. 

Figs. 136-138, while in others the extra metal is slabbed off on 
the forward and after sides until the thickness in the fore and 
aft direction is uniform from top to bottom. 

As noted in [6], the connecting rod, in the most advanced 
practice, is often hollow, a hole of uniform bore being drilled 
from one end to the other, as shown in Fig. 137. 

[9] Crank Shafts. 

Modern marine crank shafts are of two principal types, 
forged and built up. Fig. 139 shows a portion of a built-up 
crank shaft. It consists, as shown, of two crank urbs or 
throws, A and B, one crank-pin C and two portions of shaft D, 
E. Built-up crank-shafts are usually made contiuous for the 
whole engine, and in such case the piece of shafting D connects 



the crank shown with the one next to it, and thus serves as a 
common member for the two. 

In this type of crank-shaft the various sections of shaft, the 
crank-pin and the webs, are all made separately, and then fitted 
and secured together. This is usually done by shrinking and 
keying the various cylindrical members into the sections of web 
as shown in the figure. 

Fig. 140 shows a section of a forged crank-shaft. In this 
case a- forging of suitable form is made, and the various parts 
are then formed by cutting out and machining this forging. In 
many cases, moreover, the series of such sections for the entire 
engine are forged and machined in one piece, the result being a 
continuous forged crank-shaft. In other cases the section for 
each crank, as shown in the figure, is forged and made sep- 

Fig. 139. Section of Built Up Crank Shaft. 

arately, the various sections being then secured together by 
flange couplings, as shown in Figs. 140, 141. The advantage 
of making the shaft in sections lies in the fact that in many cases 
the sections may be made interchangeable, and thus a single 
spare section is sufficient for the replacement of any section 
which may become disabled through accident, and in any event 
a break will usually require the refitting of a single new section 
instead of an entire shaft. 

Forged crank-shafts are commonly used in naval practice, 
and in general where the type of construction is of specially- 
high grade and the saving of weight an important feature. 
Their use in all departments of marine practice seems, more- 
over, to be on the increase. Built-up crank-shafts, however, are 
still much used in the mercantile marine, especially where the 



conditions are easily fulfilled, and their somewhat greater 
weight is not a serious objection. 

As noted in [6], the cylindrical members of marine engines 
are often made hollow, especially in the more advanced types of 
design. Fig. 140 shows a crank-shaft section with hollow pin 
and shaft. As this feature was first commonly introduced in 
connection with shafting, and is more often met with here than 
elsewhere, the advantages of such construction may be now 

The advantages of a hollow cylindrical member such as a 
piston-rod, connecting rod, crank-pin, or length of shafting, are 
two in number, (i) It is stronger for a given weight, or for a 
given strength less weight is required. (2) The central core of 
metal is removed, and this is the most liable to contain cracks 
or flaws, which might in time extend out into the remaining 

Marine Engi>-Mring " 

Fig. 140. Section of Forged Crank Shaft. 

metal, and thus seriously weaken the member. Furthermore, 
the hole gives opportunity for the inspection of the metal on the 
inside, and thus increases the opportunity for the detection of a 
flaw which might not extend to the outer surface, or which 
might there be so small as to be overlooked. 

For cross-breaking or for torsion the metal in the ii-- 
terior of a cylindrical member is of comparatively small value. 
Thus in a lo-inch shaft, the inner core 5 inches in diameter is 
worth no more than a shell of metal about .16 inch thickness 
lying next the outer surface. Or, as a further illustration, a 16- 
inch shaft with a lo-inch hole is equal to a 1 5-inch solid shaft. 
In other words, a shell of metal 1-2 inch in thickness all around 
added on the outside of the 1 5-inch shaft will make up for the 
removal of the inner core of 10 inches diameter. In the latter 
case the hollow shaft would weigh about 65 per cent of the 
equivalent solid shaft. The saving in weight for a desired 



strength may thus be very considerable, but it is probable that 
the advantages noted above under (2) are of still greater im- 
portance, and in some cases might justify the added cost of 
making the member hollow where such addition could not be 
justified by the saving of weight only. 

Fig. 141. Detail of Flange Coupling and Bolt. 

[10] I/ine Thrust and Propeller Shafts. 

From the crank-shaft the motion is carried on to the pro- 
peller by means of a number of lengths or sections of shafting 
according to the distance from the engine to the stern of the 
ship. Of these sections the one to which the propeller is at- 
tached is known as the propeller shaft. One length must also 


Fig. 142. Flexible Coupling. 

be specially fitted to transmit the forward thrust to the thrust 
bearing and thence to the ship. This section is known as the 
thrust shaft. Other intermediate lengths form the line shafting. 
These various lengths, with the exception noted below, are 
usually connected by flange couplings, as in Figs. 140, 141. The 
coupling from the engine to the next section of shafting aft is 



often made of such form as to allow a certain degree of flexi- 
bility between the line shafting and the crank shaft. A form of 
such coupling is shown in Fig. 142. One of the coupling flanges 
is faced off, as shown like the segment of a sphere, with a ball 
and socket joint at the center to keep the two parts in line. The 
coupling bolts are then set up with nuts bearing on some form 
of spring washer which will take up the slack as the shaft re- 
volves, even when not exactly in line. The action of the coup- 
ling will be readily seen from a study of the figure. Various 
other styles of coupling are in use, but the one shown will suffi- 
ciently illustrate the principles involved. 

The thrust shaft will more naturally find its description 
with the thrust bearing. See [ 1 1 ] . 

The propeller shaft is formed with the after end tapered 

Marine Enyinetring 

Fig. 143. Outboard Shaft for Twin Screw Ship. 

and fitted with screw thread for a nut, as shown in Fig. I43C. 
The propeller is fitted with a corresponding taper, and is held 
in place by a nut and prevented from turning on the shaft by one 
or more keys, as indicated in the figure. 

In the case of twin screws, where the propeller shafts pass 
outside the skin of the ship some distance forward of the stern 
it often becomes necessary to form the outboard shaft in more 
than one length, coupled together by flange couplings as above 
described. In all cases it is necessary to form one end of the 
section of shaft which passes through the skin of the ship with 
a plain end, so that it can be passed through 'the outboard bear- 
ing as described in [n]. In cases therefore where the propeller 


cannot be attached directly to the plain end, as in single screw 
ships, it becomes necessary to provide a special form of socket 
coupling connecting the after end of the last inboard length of 
shafting with the forward end of the length which passes 
through the ship. The general plan of this arrangement is 
shown in Fig. I43b, and the details of the coupling in Fig. 144. 
Such couplings vary somewhat in detail, but the form shown in 

'Marine Enginetring 


Fig. 144. Detail of Socket Coupling. 

the figure will serve to illustrate the type. It consists, as shown, 
of the enlarged end of the inboard shaft, in which is bored out 
a tapering socket of appropriate size to take the tapered for- 
ward end of the first outboard length of shaft. The two are 
then secured together by keys and locking ring, as shown in the 

[11] Bearings. 

The various types and forms of bearing and bearing surface 
to be found in a marine engine may be conveniently examined 
under one heading. 

(i) Crosshead and Guides. The stationary part of this bear- 
ing has been already referred to in [2], and as there noted is 
usually of a hard and fine grained cast iron. The moving sur- 
face on the crosshead is usually of brass, bearing-bronze or 
white metal. When of brass or bronze it is in the form of a 
bearing piece secured to the crosshead, as shown in Fig. 130. 
When of white metal a suitable slab of brass, cast i-on or cast 
steel, with shallow pockets formed in its surface, forms the bear- 
ing piece. These pockets have slightly overhanging edges, and 
into them molten white metal is run, the general layout and 
arrangement being similar to that for the main pillow block 
bearings shown in Fig. 146. The white metal is then machined 
down to a bearing surface, in some cases being hammered with 
a round pene hammer in order to compress or harden the metal. 



The spaces between the pockets thus become spaces between 
the sections of white metal, and serve for the circulation and 
supply of oil to all parts of the bearing surface. In some cases 
shallow channels or oil grooves are cut in the guide surface or 
stationary part as well, but this is the least necessary with the 
arrangement of white metal as described. 

Liners or packing pieces are often placed between the bear- 
ing piece and the crosshead, so as to allow for adjustment and 
take up in case of wear. 

(2) Crosshead Pins. The general arrangement for the 
bearings are indicated in Figs. 133, 136, 137. The crosshead 

Marine Engineering 

Fig. 145. Main Pillow Block, Cap. 


pins are steel, and the bearing surface is brass, bronze or white 
metal. The bearing pieces are two in number, forming between 
them the hollow cylindrical bearing, and are held in place by 
steel caps as shown. From the fact that in former practice such 
bearing pieces were almost universally made of some grade of 
brass, they are still usually known as brasses. 

When white metal is used it may be fitted in the same gen- 
eral manner as above described for the crosshead slides, or in 
some cases, especially for small surfaces, the white metal sec- 
tions are turned down until a continuous bearing surface is ob- 
tained on both white metal and brass. For the distribution of 



oil, grooves or channels are then cut to serve in place of the 
channels between the sections of white metal, as in Fig. 146. 

(3) Crank Pin. The usual arrangement of this bearing is 
sufficiently shown in Figs. 136, 137. In modern practice the 
material used is commonly white metal in a brass backing or 
bearing piece, as already described. In older practice brass or 
bronze was commonly employed. 

(4) Pillow Block or Crank Shaft Bearings. The usual ar- 
rangements are shown in Figs. 120, 145, 146. Here likewise 
in modern practice the usual surface is white metal in a brass 
backing piece. In all bearings for cylindrical elements, as the 




Fig. 146. Main Pillow Block Bearing. 

crosshead pin, crank-pin, crank-shaft, etc., the two brasses are 
held from separating by a cap and bolts as shown, and from 
pinching the pin or shaft too tightly by filling pieces or liners. 
These by adjustment allow of take up for wear. In modern 
practice the lower pillow block brass is often made in the form 
of a half cylinder, as shown in Fig. 146, so that by the removal 
of the cap and upper brass the lower one may be slid around 
the shaft and so removed for adjustment or repair, or a new 
brass replaced without disconnecting the crank shaft and lift- 
ing it from its bearings, as is necessary when the lower brasses 
are of the shape shown in Fig. 120. 



(5) Line-Shaft or Spring Bearings. In these bearings the 
chief load is the weight of the shaft, at least so long as the bear- 
ings and shaft are in line and adjustment. It is therefore quite 
common to provide for such bearings simply a lower brass or 
bearing piece, usually in modern practice of white metal backed 
by brass, or in some cases by cast iron or steel. A bearing cap 
or cover is then fitted, not in contact with the shaft, and serving 

Fig. 147. Plain or Spring Bearing. 

simply to protect the bearing surface and to support grease 
cups or other lubricating arrangement. A bearing of this type 
is shown in Fig. 147. 

(6) Thrust-Bearing. At this bearing the thrust coming 
from the propeller is taken off the shaft and transferred to the 
ship. The length of shafting which is specially fitted for this 
purpose is known as the thrust-shaft. The special provision on 
the thrust-shaft consists of a series of rings or collars, as shown 



in Figs. 148-150, while the bearing of the type shown in Fig. 149 
has a corresponding series of channels into which the shaft rings 
enter when the thrust-shaft is in place. The bearing thus comes 
on the forward faces of the shaft rings and after faces of the in- 
termediate bearing rings when the propeller is turning ahead, 
and vice versa when backing. The faces of the bearing rings 
are usually of white metal, thus giving a steel on white metal 
pair of surfaces. In order to take the weight of the thrust-shaft, 
a support of brass or white metal of the usual spring bearing 
form is usually provided at the forward and after ends of the 
bearing casing. The casing is furthermore commonly made in 
the form of a rectangular box, so that it can be filled with oil and 
thus flood the bearing with lubricant. Where the shaft passes 
through the ends of the casing, a stuffing-box or form of pack- 
ing ring is provided to prevent the oil from leaking through. At 
the bottom of the casing a hollow space is often provided con- 
necting freely with the general oil receptacle above, and through 





-) (- ' 






ij a ~* 






Marine Engineering 

Fig. 148. Thrust Shaft. 

which a pipe of copper or thin brass is led back and forth. 
Through this pipe cold water is circulated for cooling down the 
oil, and thus absorbing the heat of the bearing. The base of the 
bearing is secured to the ship through a seating specially 
strengthened and stiffened to take the thrust from the shaft and 
thus transfer it to the structure of the ship. 

In Fig. 150 is shown a bearing somewhat more modern in 
type and very commonly met with in present day practice. It is 
known as the horseshoe collar bearing. The shaft is fitted the 
same as in Fig. 149, but the bearing, instead of being fitted with 
series of fixed rings and intermediate channels, is provided with 
a series of separate collars of the form shown in Fig. 151. These 
collars are provided with ears or lugs A and B, by means of 
which they are carried on side rods attached to the bearing 
casing, as shown in Fig. 150. These lugs in turn bear against 
adjusting nuts on the side rods as shown. In operation the 
thrust is transferred from the shaft rings to the faces of the col- 





lars, thence through the lugs and nuts to the side rods, thence 
to the bearing casing, and thence to the ship. 

It is readily seen that this arrangement allows of the indi- 
vidual adjustment of each collar as may be required by wear, 
or, if need be, of its removal and replacement by a spare collar 
even when under way, and without interfering with the action of 
the other parts of the bearing. 

The collars are usually of cast steel, or, in some cases, of 
brass or bronze, and the bearing surface of white metal, carried 
in pockets, as explained above. 

Marine ngiiieerinf 

Fig. 151. Detail of Collar Thrust Bearing. 

The arrangement of the casing as a receptacle for oil, the 
provision for spring bearings at the ends, and the provision of 
circulating pipes for cooling water, are similar to those above 
described in connection with Fig. 149. 

The thrust bearing is variously located. In some cases it is 
placed immediately aft of the engine with its base connected 
directly to the engine bed-plate. In other cases it is placed at 
the after end of the inboard shaft, or just forward of the after 
shaft-alley bulkhead, and in others at some intermediate point. 

(7) Stern Bearing, The general arrangement of the stern 


1 37 

bearing for a single screw ship is shown in Fig. 152. The en- 
tire distance through the stern of the ship from the stern-post 
to the after bulkhead of the shaft alley is lined with a tube A B, 
usually of cast iron. Within this are placed brass tubes C D, 
each perhaps about one-third the total length. These bearing 

tubes are provided with longitudinal 
channels slightly dovetailed in sec- 
tion, as shown in the figure, and into 
these are forced blocks of lignum 
vitae for a bearing. The arrange- 
ment is therefore somewhat similar 
to that for white metal, as described 
above, except that lignum vitae is 
substituted for white metal, and is 
placed in continuous channels run- 
ning the entire length of the bearing 
tube with intermediate spaces be- 
tween. The shaft itself is cased with 
a brass sleeve or casing, so that the 
bearing surfaces are brass on lignum 
vitae. It is found by experience that 
water is a lubricant for such a pair of 
surfaces, and it is chiefly for this rea- 
son that lignum vitae is so common- 
ly used as the material for the sta- 
tionary part of the bearing. The 
brass sleeve, which preferably ex- 
tends the whole length of that part of 
the shaft within the tube, also pro- 
tects the shaft from corrosion. This 
part of the shaft is known either as 
the tail shaft or propeller shaft. It is 
usually made a little larger than the 
line shaft to provide for corrosion, 
and also for the more violent shocks 
to which it is subject. At the forward 
end of the tube A is fitted a stuffing- 
box, through which the shaft passes 
to the after end of the shaft alley. 
At the after end of the tube the 
water enters freely through the 



spaces between the lignum vitae and flows forward, thus serv- 
ing to cool and lubricate the bearing. At the forward end the 
stuffing-box prevents leakage through into the ship. It is de- 
sirable, however, to fit a small pipe and cock so that water 
may be drawn from the tube as desired, in order to judge by 
its temperature as to the condition of the bearing. Instead of 
a pipe and cock the stuffing-box follower is sometimes loos- 
ened up so as to allow a sufficient leakage to insure circula- 
tion through the tube, and to serve as an index of the condi- 
tion of the bearing. 

In some cases, instead of water lubrication, the after end of 
the stern tube is closed against the water, and the tube is filled 

Fig. 153. Stern Brackets. 

with heavy oil or tallow. Or. if desired a stand pipe may be run 
up to a sufficient height, so that when filled with oil it will pro- 
duce a pressure in the tube slightly greater than that of the 
water outside, and thus the leakage will be outward rather than 
inward. With oil lubrication the lignum vitae bearing surface is 
replaced by white metal. 

In small craft the steel shaft without casing is often fitted 
directly in a brass bushing or bearing. In such case oil lubri- 
cation is to be preferred, but very commonly the bearing is left 
to run with such lubrication as the water can provide. 

For twin screw ships, as shown in Fig. 143, the same gen- 
eral arrangement 'is used, except that the length of the tube 



where it passes through the skin of the ship is shorter, and fre- 
quently the lignum \itae bearing extends the entire length in- 
stead of over a part of the forward and after ends. A similar 
form of bearing is also provided in the shaft brackets or struts 
just forward of the stern post. The general form of such brack- 


Fig. 154. Stern Bracket Bearing. 

ets is shown in Fig. 153. On each side is a heavy steel casting 
secured firmly at top and bottom to the structure of the ship, 
and carrying at the apex a boss for the bearing, as shown in 
Fig. 154. This bearing is formed by a tube carrying lignum 
vitae strips as previously described, and in this way, with twin 
screws, the extreme after ends of the shafts are supported. 


The peculiar conditions existing on the western rivers of 
the United States have resulted in the development of a special 
type of boat and propelling machinery. In the early days of 
river navigation the raft was first employed, and then came the 
flatboat, which has stood as the type of all later developments. 
On the rivers where at certain seasons of the year the water is 
shallow, the current swift and the channel narrow and tortuous, 
the usual style of keel boat would be of small service, while the 
light draft flat bottom craft seems admirably adapted for navi- 
gation under such difficulties. 



Of the two varieties of boat, side wheel and stern wheel, 
the latter is preferred as on the whole the better suited to the 
all-around conditions of river navigation, and the flat-bottomed 
stern wheel craft may to-day be considered as the typical boat 
for western river navigation. Indeed this type of boat has met 

*Vrwte Engineering 

Fig. 155. Western River Engine, Elevation and Section through Valve Chests. 

with much favor for river navigation in all parts of the world, and 
especially in South America, where they are largely employed. 
The type of engine used on western river boats is shown in 
Figs. 155-157. It is horizontal and of the simple non-con- 
densing type. Two such engines are usually employed, .one on 
each side placed close to the guards, with the axis of the cylin- 

Fig. 156. Western River Engine. Top View. 

ders fore and aft, and with the connecting-rods coupled to the 
cranks on the stern wheel paddle shaft. 

The cylinders are of relatively small diameter and long 
stroke, the dimensions in a typical case being 24-inch diameter 
by 96-inch stroke. The most peculiar feature of the engine, 
however, is found in the valve gear. The valves themselves are 
usually of the double-beat poppet form, as shown in Fig. 155, 



and each cylinder is provided with four, two for steam and two 
for exhaust. These valves are actuated by a cam valve gear 
mechanism, as briefly described below. 

The steam valves with their connecting pipes are located 
on one side of the cylinder, while the exhaust valves and con- 
nections are on the other side. Each set of valves is operated 
by separate rocking cams or levers, which receive their motion 
through rockers and connections from a special cam located on 
the main paddle shaft. 

The cam type of valve gear possesses peculiar advantages, 
especially for long stroke, slow revolution engines such as are 
used in these cases. The motion of the valve may thus be made 
intermittent, giving a quick opening and closure, with interme- 

Marine Enyliieering 

Fig. 157. Western River Engine, End View and Section Showing Valves. 

diate periods of rest or very slow motion. It is also peculiarly 
adapted to the elastic movements of the boat during the process 
of loading, unloading, etc., movements which continually vary 
the distance between the main shaft and rock shaft, and which, 
with almost any other type of gear, would introduce serious dis- 
turbance into the movement of the valve and the distribution of 
the steam. 

For the operation of these valves in the common type of 
gear two cams are used ; one known as the full stroke cam and 
one as the cut-off cam. When the engine is in full gear the full 
stroke cam operates all four valves, raising one exhaust and 
one receiving valve at opposite ends of the cylinder at the same 
moment, and alternately at each end, thus distributing the steam 



as required to carry the piston back and forth continuously. 
The one cam does all the work in the full gear motion of the 
engine both ahead and astern, and is hence in its neutral posi- 
tion when the crank is at its dead point. The cut-off cam is so 
arranged as to be hooked on after the full stroke cam has given 

Fig. 158. Full Stroke Cam with Yoke. 

headway to the boat, and is used in the go-ahead motion only. 
This cam is so designed that the steam is cut off at any desig- 
nated point in the stroke, as at J^, JH$, 3^, etc. 

The form of a full stroke cam with its yoke is shown in 
Fig. 158, and of a ^ stroke cut-off cam in partly dotted lines. 

Fig. 159. Full Stroke Cam. 

In Fig. 159 is shown the usual type of construction of the full 
stroke cam, and in Fig. 160 similarly the 3,4 cut-off cam. 

With this arrangement of gear the exhaust is opened and 
closed just at the end of the stroke, and hence neither early ex- 
haust opening nor closure for cushion can be obtained. A 



means of obtaining the former has been found by blocking up 
the exhaust lifters somewhat, so that the valve will be slightly 
open when the engine is on the dead point. 

This insures an earlier opening of the exhaust and so clears 
the cylinder for the return stroke, but it gives likewise a later 
exhaust closure, so that with the engine on the center both ex- 
haust valves are slightly open, and in full gear operation a slight 
"blow through" will occur. This disappears, however, when the 
cut-off cam is engaged, because the opening movement of the 
latter is much slower than that of the full stroke cam. 

Various modifications of this simple cam gear have been 
introduced with a view of improving the general operation, 
especially by the provision of means for obtaining both steam 
and exhaust lead and compression, as well as independent move- 
ments for the go-ahead and backing motions. 

Fig. 160. Three Quarter Cut-off Cam. 

In the Sweeney valve gear two full stroke cams are em- 
ployed, one for go ahead and one for backing, each set so as to 
give suitable exhaust lead and compression, while a separate 
cut-off cam is fitted for the go-ahead motion. 

The crossheads of these engines are usually of the locomo- 
tive type, with long brass gibs bearing on the top and bottom 
guides. The connecting-rods are commonly of wrought iron or 
wood, with iron or steel fittings, and form one of the most pe- 
culiar features of these engines. Wood is often thus preferred 
over metal because it seems to be better capable of standing the 
shocks and peculiar twisting strains which come upon the rod, 
and in spite of the strangeness of the combination, we find in 
some modern boats a fluid compressed nickel steel paddle shaft 
with a wooden connecting-rod. The rods are very long, fre- 
quently as much as eight times the crank, and the best rods are 


made of Oregon fir, reinforced with iron straps which are let 
into the body of the rod and through bolted. The ends of the 
rods are fitted with brass boxes with straps, gibs, keys, etc., in 
the usual manner of fitting up such form of rod, and as illus- 
trated in Fig. 138. 

In some cases of modern river boats on the Pacific Coast 
many changes have been introduced looking toward a closer 
approach to usual marine practice. In a typical example of such 
improved practice the engines are horizontal tandem compound, 
the high-pressure cylinder having piston valves and the low- 
pressure cylinder slide valves, both operated by excentrics and 
link work in the usual way. In this case there are two engines 
developing about 1,500 I. H. P. each. The cylinders are 22^/2 in. 
and 38^4 in. diameter, with a stroke of 8 feet, and are intended 
to make thirty revolutions per minute. 

The crank shaft for such engines is built up in structure, 
the two cranks being separately forged and secured to the pad- 
dle shaft by shrinking and appropriate keys. The shaft is 
usually fitted with hexagonal bosses where the wheel flanges 
are to be secured. The latter are usually of cast iron, heavily 
ribbed and reinforced by wrought iron bands shrunk on their 
hubs and outer circumference. These flanges are fitted to the 
hexagonal bosses on the shaft, and are secured with suitable 
keys. They are provided on one face with sockets for the wheel 
arms, which are of wood. These latter are further strengthened 
by circular bands of iron bolted near the outer ends, and also by 
oblique bracing which is worked between them. 

The buckets are also of wood, 2-inch oak plank of suitable 
width and length being a standard material. They are secured 
to the wheel arms by special clamp bolts, and are so located 
relative to the draft of the boat as to be immersed only some 
4 to 6 inches when the steamer is running light. In some cases 
the buckets are divided at the center, forming really two sets, 
staggered with reference to each other, and thus reducing the 
shock of the wheel as it enters the water. 


This peculiar feature of western river practice as illus- 
trated in Fig. 161, is a combination of feed pump and feed water 
heater. As here shown, the doctor consists of a vertical beam 
engine with crank and flywheel operating four pumps. Two of 



these are simple lift pumps drawing water from the river and 
delivering it into the heating chambers overhead, while the 
other two are feed pumps proper, taking their supply from the 
heaters and forcing the water into the main boilers. Each lift 
and force pump is designed of sufficient capacity to supply the 
entire battery of boilers, so that one of either kind may be dis- 

Marine Engineering 

Fig. 161. Western River Boat "Doctor." 

connected for examination or repair without disturbing the 
regularity of boiler feed supply. The various parts of the ma- 
chine are erected on a deep cast iron base plate which contains 
various ports and passages, forming the water connections be- 
tween the various pumps. 

The suction pipe from the river is connected with a vacuum 
chamber, and communicates through a passage in the base cast- 


ing with the suction side of the lift pump. The discharge from 
these pumps is then led by other passages to the columns, which 
serve as discharge pipes, supports for the engine beam and for 
the heaters. Valves are also located in these columns, by 
closing which the water in the heaters may be prevented from 
returning at such times as it is necessary to open up a pump for 
examination or repair. 

The heaters themselves consist of wrought iron shells riv- 
eted to cast iron heads, through which the exhaust steam from 
the main engines is led on its way to the exhaust pipe. The ex- 
haust steam thus comes in direct contact with coils of copper 
pipe that lie in the lower part of the heaters, and through w r hich 
the feed water is forced and finally discharged below a dia- 
phragm. Beyond this the exhaust steam and water are to some 
extent in direct contact, the latter being finally lead down 
through the pair of columns on the opposite side of the machine 
to the feed pump inlet valves in the base-plate. The head of 
water in the columns is thus sufficient to flood the valves and 
prevent the pump from missing stroke, even with the hottest 
feed water which the heaters can furnish. 

The lift pumps are fitted with long pistons having either 
cup leather or square gum packing, while the feed pumps are 
of the common plunger type. The pump valves are flat disks of 
brass made quite thick so as to avoid the need of springs, and 
also to allow metal for re-facing. The engine part of the doctor 
is very simple and will call for no special comment, consisting 
simply of a steam cylinder for actuating the beam and thus giv- 
ing motion to the four pumps as described. 

In some cases of recent river practice injectors have taken 
the place of the "doctor," and if they can be depended upon 
they are of course much preferable, being easy to handle and 
occupying little or no space otherwise valuable. While it is 
probable that they can thus be used to advantage on certain of 
the upper portions of the western rivers, it is hardly possible 
that they could be used at all in many other localities on ac- 
count of the sand and grit which is held in suspension by the 
water, and which would cut out the injector tubes so rapidly 
that their use would be out of the question. For this reason it 
seems likely that the "doctor" will hold its own in all such locali- 
ties and that it will continue to be an important detail of west- 
ern river practice. 




Within the past few years the application of the steam tur- 
bine to marine propulsion has produced results which have at- 
tracted world-wide attention, and it seems at present not un- 
likely that the part taken in the future developments of marine 

propulsion b> this form of mo- 
tor will be one of increasing im- 
portance. The present develop- 
ment is represented by the Par- 
sons' steam turbine as fitted on 
the Turbinia, and later on the 
British torpedo-boat destroyers 
Cobra and Viper, and Clyde 
passenger steamer King Ed- 

This type of motor, one form 
of which is represented in Fig. 
162, consists of a cylindrical 
case carrying rings of inwardly 
projecting oblique guide blades, 
while within these revolves a 
shaft carrying rings of outward- 
ly projecting oblique blades. 
There is a clearance of about 
1-16 inch between the successive 
rings of blades and guides, and 
between the ends of the former 
and the case, and the ends of the 
latter and the body of the shaft. 
There is thus left between the 
shaft and the case an annular 
space filled with alternate rings 
of blades attached to the shaft 
and guides attached to the case. 
The steam when admitted passes 
first through a ring of fixed 
guides by means of which it is 
given a rotational motion, and 
then projected on to the first 
ring of blades. It is then thrown 
on to the following ring of 


guides, by means of which its direction is again changed, and it 
is then thrown in the same direction as at first upon the second 
ring of blades, and so on, passing from one ring to the next, 
and giving to each a part of its energy and thus maintaining 
the rotation of the shaft. It must be especially understood that 
the steam in a turbine acts by impact or reaction, and not by 
pressure. As the steam rushes against the blades and its direc- 
tion of flow becomes thereby changed, it exerts a reaction on 
the blades, and this constitutes the force which produces the 
rotation. The energy which the steam possesses in virtue of its 
temperature and pressure cannot therefore be directly used in 
the turbine as in the common steam engine, but it must first be 
transformed into the energy of motion as in the rushing jet. 
This transformation is effected by the gradual expansion of the 
steam as it passes through the turbine. This gradual expansion 
keeps up a continual transformation of heat energy into the 
energy of motion, and this energy is as constantly transferred 
to the moving blades, and thus the original heat energy of the 
steam is transformed into mechanical work. In order that the 
expansion may be carried on continuously, the cross-sectional 
area occupied by the blades is increased from time to time, as 
shown in the figure, thus making a series of steps of increasing 
diameters for the rings of blades and guides. In a moderate 
sized motor there may be 50 to 80 successive rings thus ar- 
ranged in from three to five groups. 

The steam pressure acting on the annular rings separating 
these steps, and also on the blades themselves, would set up a 
severe end thrust. This may be balanced, as in Fig. 162, by ar- 
ranging two sets of such steps on the same shaft varying in op- 
posite directions, and thus balancing each other in end thrust. 

In another method special disks are fitted on the shaft with 
grooves in their outer surface, within which project rings car- 
ried in the frame. These rings and grooves form a nearly steam 
tight joint, and the annular area is so adjusted in amount that 
the steam acting on it will balance the thrust acting on the 
blades and corresponding annular area of the main shaft. In 
addition a special thrust bearing is fitted for taking any residual 
end thrust and for making the necessary adjustments. 

The arrangement thus briefly described constitutes a sim- 
ple turbine. By combining two or three such in series and 
leading the steam from one to another through them, the so- 


called compound turbine is formed. In a simple turbine the 
steam may be expanded some six or eight times, so that by 
compounding a total expansion of, say, 30 to 60 could be ob- 
tained, and by a triple combination a total expansion of 200 and 
more may be effected. 

The revolutions of the turbine are high. This is a necessity 
of its efficient performance. The Turbinia has three each of 
about 600 to 700 horse power, and the revolutions run up to 
about 2,000 per minute. In the Viper and Cobra there are four 
turbines, each of about 1,500 H. P., running at about 1,200 
revolutions per minute. With the former a speed of about 34 
knots has been reached, while with the latter speeds of over 35 
knots were reached. 

The efficiency of the compound type of turbine is not very 
different from that of good compound or triple expansion en- 
gines of the usual type. The weight for the same power is 
somewhat less, though the difference is not great when com- 
pared with light reciprocating engines forced in a manner cor- 
responding to the conditions on these boats. In other words, 
the turbine shows no very great saving over the weights of the 
lightest types of reciprocating engines as designed for torpedo 
craft ancl fast launches. 

The great advantages of the turbine are found in its com- 
pactness and absence of reciprocating motions, and thus in the 
entire absence of the forces which cause vibrations. With ex- 
treme speeds and in many conditions of modern practice the 
vibration forces become a serious question, and a motor en- 
tirely free from them becomes immediately of great importance 
in all such cases. 

In regard to questions of durability, maintenance, liability 
to derangement or accident, etc., the experience available at 
present is too small, and we must wait for the future to answer 
these and many other questions which bear on the applicability 
of the turbine to the various conditions of marine practice. 

[i] Throttle Valve. 

The purpose of the throttle valve is to provide a means for 
quickly opening or closing the main steam-pipe near where it 
connects with the high pressure valve chest, and thus to provide 
for the quick control of steam to the engine when stopping and 



starting. A great variety of valves have been employed for this 
purpose. The necessity for quick operation, especially by hand 
gear, requires usually some form of balanced valve, though in 
very small sizes an ordinary globe or straight-way or gate valve, 
as shown in Figs. 166, 168 may be used. Of these the straight- 
way valve is much to be preferred, as when open it leaves prac- 
tically an unobstructed passage for the flow of the steam. 

(i) Gridiron Valve. The gridiron is another form of unbal- 
anced valve sometimes employed as a throttle. This valve, as 
shown in Fig. 163, consists of a series of bars and ports corre- 
sponding to a like series in the valve chest, and giving a series of 
openings for the steam, wider or narrower according to the 
position of the valve. With such an arrangement a considerable 
area of opening may be obtained with a comparatively small 

Fig. 163. Gridiron Valve. 

Fig. 164. Double Beat Poppet Valve. 

movement of the valve, and a screw or some other form of slow 
motion gear may be employed without loss of quick opening 
and closure. This form of valve is, however, but rarely met with 
in modern. practice. 

(2) Double Beat Poppet. The double beat poppet valve, as 
shown in Fig. 164, has been much employed as a form of bal- 
anced throttle. The upper disc is slightly larger in area than the 
lower, so that if the live steam is on the outside the net load on 
the valve is that due to the difference of the two areas, and this 
may be made very small. The resistance to opening is thus no 
more than can be readily overcome with a direct hand gear, as 
for example, a simple lever or other like arrangement. 

The chief difficulty with this valve is in keeping it tight, var- 
iations of temperature and the consequent expansions and con- 
tractions often tending to slightly unseat one disc or the other. 

(3) Butterfly Valve. The butterfly valve has also been wide- 



ly used as a balanced throttle. It consists of a disc of elliptical 
form carried on a spindle and swinging within a cylindrical 
casing. When closed it rests obliquely on the inner surface of 
the casing, thus closing the passage around its outer circumfer- 
ence. When full open it swings into a position with its plane 
lying along the pipe, thus leaving the passage nearly free for the 
flow of steam. This form of valve is quite perfectly balanced, 
but it is difficult to keep tight. If the angle of obliquity with the 
surface of the casing is too small, it may also be liable to stick 
fast, due to unequal expansion of the valve and casing. In an- 
other form of butterfly valve, as shown in Fig. 165, however, the 

Marine Enginitrmg 

Fig. 165. Combined Stop and Throttle with Balance Piston. 

disc is circular and when closed swings square across the line of 
flow, just fitting within a corresponding ridge of the casing. In 
such case the diameter of the opening must be made enough 
larger than that of the disc to avoid the danger of striking, and 
considerable leakage will usually result. 

(4) Disc Valve With Balance Piston. A plain disc with bal- 
ance piston attached to the stem is quite commonly employed in 
modern practice for the throttle or for the stop and throttle com- 
bined. Such an arrangement in combination with a butterfly 
valve is shown in Fig. 165. By this means the pressure on the 
piston nearly balances the load on the valve, and it may thus be 
operated by hand gear. Steam may also be admitted back of 
the piston by a pipe with stop-valve operated from the working 
platform. By this means the disc may be balanced when once 
off the seat, and closure effected as easily as opening. 



(5) Power Operated Throttle. In some cases 1 with large en- 
gines the throttle is operated by steam power instead of by hand, 
steam being admitted to an operating cylinder by means of a 
hand lever or other like arrangement. Here the steam acts 
upon an auxiliary piston and by suitable connections produces 
the movement of the throttle as desired. In such cases the con- 
nections are often of the "floating lever" type, as in the reversing 

Fig. 166. Globe Valve. 

gear described in [5,] so that the valve will follow the hand lever 
in its movement back and forth, and the combination becomes 
thus equivalent to a direct operation of the throttle by hand. 

[a] Main Stop Valve. 

The throttle valve from its construction can rarely be closed 
sufficiently tight to prevent leakage of steam, often considerable 
in amount. To provide a shut-off without sensible leakage a 



stop valve is often fitted in addition. Such valves may be of 
various types, as shown in Figs. 166, 167, 168. 

(i) Globe Valve ^ This valve, as shown in Fig. 166, consists 
of a metal chamber of globular or spherical form with flanges for 
connecting to the line of piping. Within the body is a partition 
separating the portions connected with the two openings, and in 
this partition is a hole with conical seat upon which the valve 
with corresponding conical face bottoms when closed. The 
valve is attached to a threaded spindle which works in a nut 
either formed in the neck which contains the stem, or carried 

Marine Engineering 

Fig. 167. Angle Stop Valve. 

outside on a girder supported by stud bolts, as shown in the fig- 
ure. To the end of the stem a handle is attached, and by this 
means the valve is opened or closed as desired. The stem is 
packed by means of a stuffing-box and soft packing compressed 
by a gland of the usual form as shown. In small sizes the gland 
is usually replaced by a form of nut threaded to the neck, which 
contains the stem, and compressing the packing between the nut 
and the bottom of the packing space. 

(2) Angle Valve. In this type of valve, which is an angle 
or elbow and a valve combined, the seat and valve-face, as shown 



in Fig. 167, are placed square across one of the openings, thus 
shutting off all flow through it when the valve is closed. When 
the valve is opened, however, the passage is left free, according 
to the degree of opening, for the flow of the liquid or vapor 
around the angle and on into the following section of pipe. 

When the stop valve is of the disc form it is very commonly 
of the angle type and arranged to go in at a turn of the pipe, as 
shown in Fig. 167. In this case also the valve is attached to a 
bulkhead and the arrangement will serve to show the method of 

Marine Engineering 


Fig. 168. Gate Valve. 

carrying steam through a bulkhead and of making up the joints 
connecting together the steam pipe, the stop valve and the bulk- 
head plate. 

(3) Straight-way or Gate Valves. In this form of valve, 
which is shown in Fig. 168, the moving part consists of a special 
form of slide which is moved by a screw back and forth across 
the opening of the pipe. There are various special forms and de- 
vices for securing tight contact between the valve and its seat 
when closed, and thus making the valve tight under steam pres- 


sure. The general arrangement of Fig. 168 will, however, serve 
to show the main features of valves of this type. When closed 
and with pressure on one side of the slide only, there is some- 
times some difficulty in opening the valve. To relieve this con- 
dition a small by-pass, as shown, is often fitted. This admits 
steam to the farther side of the valve, thus balancing the load 
and making the operation of opening much easier. In another 
form the valve slide is made of two parts, hinged together and 
with the end of the spindle working between them in such way 
that when screwed hard down it is forced as a wedge between the 
two parts thus forcing them against their seats. When the 
handle is turned in the reverse way the first action is to partly 
withdraw the stem from between the two parts of the slide, thus 
easing them from their seats and allowing them to be readily 
withdrawn as the stem is turned farther back. 

When the throttle takes the form of a plain disc with bal- 
ance piston, as in Fig. 165, no additional stop is thought neces- 
sary, and such an arrangement is often known as a combined 
stop and throttle valve. In such case, however, a screw stem 
may be provided with connections for bringing it into use when 
closing the valve down as a stop. 

[3] Cylinder Drain Gear and Relief Valves. 

A certain amount of water is likely to collect in the steam 
chests and cylinders, either carried in with, or condensed from 
the entering steam, especially when warming up the engine pre- 
paratory to getting under way. Provision must be made for 
getting rid of this water as occasion may require, and to this end 
the so-called cylinder drains and relief valves are fitted. The 
drains are usually plain cocks piped up and connected to the 
parts to be drained, and with the valve stems connected by 
levers and bell-cranks to operating handles at the starting plat- 
form. The drains in the bottom of the cylinder or valve-chest 
will naturally be placed at the lowest point at which water can 
collect, or as near to such point as is practicable. Those in the 
upper end of the cylinder will be placed at such a height that the 
opening will not be covered by the piston when at the top of 
the stroke. 

For small engines, auxiliaries, pumps, etc., the drain valves 
are often plain globe valves piped into the cylinder at convenient 
points, and operated independently by hand. 


The discharge of the drains is piped away either into the 
bilge, or into a fresh water collecting tank. 

In addition to such gear, which is operated by hand, and 
when judgment may calMor its use, it is necessary to provide 
automatic relief valves for the discharge of water in larger quan- 
tities should it find its way into the cylinder by priming or in 
other ways. Such a relief valve is in the form of a safety valve, 
and may be set to open at any pressure desired. Such valves 
are sometimes connected up with operating levers, also led to 
the starting platform, so that they may be operated by hand from 
that point. In such cases only the one set of valves is often 
fitted, automatic when necessary, and under hand control when 
desired. In some cases with large engines a double set of 
automatic relief valves is furnished, a pair of large valves not 
under hand control, and a smaller pair under hand control, as 
described above. 

[4] Starting Valves. 

In order to assist in starting the engine, especially if the 
high pressure piston happens to be on or near the center, a valve 
and pipe are usually provided for admitting steam direct from 
the steam pipe or high pressure valve chest, to the first receiver, 
or intermediate pressure valve chest. This will give sufficient 
load on the intermediate pressure piston to start the engine, and 
carry the high pressure piston off the center, and thus give the 
engine a chance to start in the regular way. In case the high 
pressure and intermediate pressure cranks should be opposite, 
and thus both pistons on or near the center at the same time, 
the auxiliary pipe will lead to the second intermediate piston, 
or to the first cylinder, whose piston is not on the center with the 
high pressure. In some cases the passage of the steam to the 
next cylinder beyond the high pressure is effected by the open- 
ing of a valve connecting the steam and exhaust sides of the 
high pressure valve chest. Such valve being opened the steam 
finds its way directly to the point where it is needed. 

Valves for this general purpose are variously called pass- 
over, or starting valves, or monkey tails. They are either in the 
form of a cock or of a small slide valve, in either case admitting 
of full opening by a single short stroke of a convenient hand 
lever, to which they are connected by suitable rods and con- 



[5] Reversing Gear. 

The various links of a Stephenson valve gear, as will be 
seen in Section 53, are connected by side or bridle rods to arms 
on the rock or "weigh" shaft. To reverse or link up with such 
a gear, therefore, it becomes necessary to provide some means 
for turning this shaft back and forth, and for holding it under 
complete control at any position desired. The form of reverse 
gear most commonly employed in American practice is of the 
so-called "floating lever" type, and is illustrated in Fig. 169. 

It consists of a cylinder, AB, with piston and rod, D, con- 
nected by a link from E to an arm on the engine rock shaft, 
and thus connecting with the links. 

Fig. 169. Floating Lever Reverse Gear. 

As the piston is moved back and forth by the steam this arm 
will evidently be carried with it, and the various Stephenson 
links, or like parts of other types of valve gear, will be moved 
as desired, each through its connection with the rock-shaft. The 
steam to the cylinder, AB, is controlled by a slide valve, V, either 
plain or of the piston type. This valve has very small lap so 
that from the position when covering both ports but slight 
movement is needed to uncover. To the stem of the valve is at- 
tached a link, LI, which at the latter point is joined to a bar, KH. 
The lower end, H, of this bar is attached to a lug, Q, on the 
piston rod, and, therefore, moves with the piston. The upper 
end, K, is connected through a link, KN, to a hand lever, which 
is provided with means for clamping in any position desired. 


Suppose now the gear in the position shown and with the 
valve covering both ports. Let the hand lever be moved so as 
to throw KN to the left. For the moment H will be a fixed 
center, and with the connections shown the valve will be moved 
to the left also. With this arrangement of connections an in- 
side valve (Section 46 [7] ) must be used, and, therefore, steam 
will be admitted to the left hand end of the cylinder, AB, and 
the piston forced to the right. Let the hand lever, carrying 
with it the valve, be thus moved over a certain distance and then 
held or clamped there, thus fixing NK. The point, K, will thus 
become for the time a fixed center, and the movement of H to 
the right will carry the valve in the same direction, and thus 
finally close the ports, shutting off the supply of steam at one end 
and closing the exhaust at the other. ' The movement of the 
piston will thus be stopped and the gear will be held in the po- 
sition reached. It is clear that for every position of the hand 
lever there will thus be some position of the piston, 
rock shaft arm and main valve gear, for which the valve will be 
brought to mid position and the piston and gear thus brought to 
rest, and that the steam will carry the gear to this position and 
then automatically shut off and stop. If the hand lever is moved 
but slightly so as to barely displace the valve, the piston will 
move but a small distance before again covering the ports antl 
coming to rest. If the handle be moved to an extreme position 
the valve will be moved far over and the steam will rush the 
piston and gear over into the extreme corresponding position. 
In short, the position of the gear for equilibrium under steam will 
correspond exactly to that of the hand lever, and wherever the 
latter may be placed, the gear will run to the corresponding 
position and then stop. It is also clear that if the hand lever be 
moved slowly, the piston and main links will follow along at 
equal pace, stopping when the handle is stopped and moving 
when it moves. Also if the handle be slightly displaced and left 
to itself the friction of moving the valve will be usually more 
than that of moving the handle, and in consequence the point, I, 
will become for the time a fixed center, and the piston will move 
along, the valve remaining open and the connection, HKN, 
moving the handle over at equal pace with the link. This will 
continue till the handle comes against a stop at the end of its 
path. The point, K, will then become fixed, and the further 
movement of the piston will move the valve into mid-position, 


thus shutting off steam and bringing the gear to rest in the 
position corresponding to that of the hand lever. 

To take up sudden shocks and provide a safeguard against 
putting the link over too rapidly and thus overrunning at the end 
through the inertia of the parts, spring stops or buffers, R, are 
provided on a rod, S, against which the lug, Q, comes at the 

end of its run. 

It is thus seen that this gear furnishes a very perfect con- 
trol over the main valve gear, the action being the same as for a 
man operating the gear directly, and thus giving him readiness 
of control with the least mental effort, and the least liability of 
error in a moment of hurry or excitement. 

In addition to the spring buffers, as shown in Fig. 169, a 
form of plunger control is sometimes added. In this arrange- 
ment the piston rod of the reverse cylinder is continued back- 
ward and connected to a second piston or plunger working in a 
cylinder filled with oil. The operation of the plunger is to trans- 
fer the oil through a suitable pipe connection from one end of 
the cylinder to the other, and as this passage may be throttled 
at will by a stop valve, all possibility of slamming or of violent 
motion may be removed. A further advantage of this arrange- 
ment lies in the fact that with suitable pipe connections to a 
hand pump, the oil may be drawn from one side of the plunger 
and forced in on the other, thus giving a control over the valve 
gear by hand power in case of derangement of the power con- 

Of other forms of reverse gear the so-called all around gear 
is quite commonly met with in English practice. The main 
links are connected up to a small engine which makes a large 
number of revolutions in running the link over from one extreme 
to the other. This engine is under the control of a small link 
which is directly operated by hand. A form of lever stop is 
usually provided which will either reverse or middle the small 
link and bring the engine to rest when the main links have 
reached either extreme of their travel. 

In engines for small yachts, launches, etc., the links are 
placed directly under the control of a hand lever. 

The various other types of valve gear may be operated by 
any of the forms of reverse described. With all valve gears, as 
described in Chapter VII, the reverse is effected by the move- 
ment of some piece of the gear from one position or location into 


another, and so back and forth for the various degrees of linking 
up, etc. By suitably connecting such piece to the power re- 
verse gear the control may, therefore, be obtained in the same 
manner as for the Stephenson link as described above. 

[6] Turning Gear. 

It is always necessary to provide some means for turning 
the engine other than by steam on the main pistons. This is 
necessary for moving the engine when in port for adjustment of 
bearings, setting of valves, etc. The turning gear usually con- 
sists of a large worm wheel placed on the main shaft just aft of 
the bed-plate, geared down through worm and spur gearing to a 
small engine, usually a double simple engine with cranks at 90. 
The gearing ratio is such that many hundred revolutions of the 
turning engine may be required to one of the turning 
wheel or main engine shaft. This gear must be so ar- 
ranged as to be readily thrown in and out of connection with 
the main turning wheel. This is usually accomplished by carry- 
ing the main worm on a shaft which is pivoted, and which can 
thus be locked in either of two positions, in one of which the 
worm is in gear, and in the other out of gear, or else by driving 
the worm on a shaft with a feather, thus providing for endwise 
motion, and for fixing it in either of two locations on its shaft, 
in one of which it is in gear and in the other out of gear. The 
latter is the arrangement more commonly met with. 

Where a turning engine is not provided the turning wheel 
is usually arranged for operation by hand through worm gear- 
ing operated by a lever with pawl and ratchet arrangement, or 
by some similar device. 

In some cases the engine is turned by a hydraulic jack 
placed under a movable chock piece located in sockets cast in 
the turning wheel. This chock is shifted from one socket to an- 
other as the jack shoves it upward, and thus the engine is slowly 

In small engines the turning wheel is often simply a form of 
gear wheel with shallow teeth in which a pinch bar is worked, 
and by this means the engine may be slowly pried around. Such 
a wheel is known as a pinch wheel. 

[7] Joints and Packing. 

The joints to be considered under this head are of two 
kinds, (i) Fixed joints as those between a cylinder or valve 


chest cover and flange, and (2) Sliding or slip joints as those be- 
tween a piston rod and the stuffing box, or the slip joint in a 
length of steam piping. 

For making up stationary joints a great variety of packings 
are in use, the difference depending to some extent upon the 
temperature to which the joint is to be subjected. Thus 
for joints to stand high temperature, as with boiler man- 
holes, cylinder heads, etc., sheet asbestos either plain or 
in combination with other materials is used. There are 
also various kinds of packing in which rubber in one form 
or another is used either in combination with some fibrous ma- 
terial as sheet canvas, or as a constituent of some form 
of compound. The tendency of rubber by itself is to grow 
dry, hard and brittle, especially under the action of heat, and the 
purpose of the modern forms of rubber compound is to avoid 
this tendency, at the same time retaining its elasticity and joint 
making qualities. For joints not subject to the action of high 
temperature, similar forms of packing are used, though with a 
greater proportion of rubber, if desired. 

The strip or ring of packing which is cut out and fitted 
for the joint is called a gasket. 

In making up such a joint it is well to smear the surfaces 
of the gasket with a mixture of black-lead and grease or oil. 
This will aid somewhat in making the joint, and very much in 
the removal of the cover and gasket at a later time without tear- 
ing the latter. With such precaution and when the temperature 
is not high the same gasket may be used several times over with- 
out loss of its joint making qualities. 

In addition to gaskets made of such materials as described 
above, joints are also made with gaskets of corrugated sheet 
copper, or of plain copper wire. For high pressures such gas- 
kets have proved quite successful. The soft copper is expanded 
between the harder metals of the flanges, and spreads, filling the 
surfaces where it touches, thus making a tight joint. 

For sliding joints as between a piston-rod and stuffing-box, 
the greatest variety of packings is likewise in use. They may 
be broadly divided, however, into the two classes, fibrous and 

The fibrous packings are made of the same material as the 
sheet packings above described, and are either round, square or 
triangular in section. For use. they are cut to such lengths as 



may be necessary and placed in the stuffing-box in layers or 
turns, the joints between the ends being shifted so as not to 
come one above another. The stuffing-box, as shown in section 
in Fig. 170, consists of a cylindrical chamber or box, EF, with 
cavity B. This is bolted by means of the flange, F, to the lower 
cylinder head. The part, CC, is known as the gland or follower 
and is carried by two or more studs, as shown. At the bottom 
or upper end of the box is a ring, as shown, just filling in the 
space between the opening in the box and the piston rod. Fre- 

Fig. 170. Plain Stuffing Box. 

quently this ring is omitted and the metal of the box fits about 
the rod. The packing is placed in the box as described above, 
thus filling the space, BB, between the bottom of the box 
and the gland. The packing may then be compressed as de- 
sired and as may be necessary by means of the nuts on the stud- 
bolts, thus forcing the gland down on the packing and making 
the joint tight. This is the general type of all such joints made 
with compressible packing, with, of course, variation in details. 
Joints of this character are used for piston and plunger 



rods, slide valve stems, globe and disc valve-stems, joints about 
the shaft where it goes through a water-tight bulkhead, joint in 
the thrust bearing casing, as noted in section 21 [n], in slip and 
expansion joints, as noted in section 25 [2], etc., etc. 

For metallic packing with joints of this character the form 
of the box is in general the same. In fact in some cases the 
box is so made that either soft or metallic packing may be used. 
Here again the greatest variety in detail is to be found, but a 

Fig. 171. Metallic Packing. 

single instance will serve to illustrate the essential features of 
such packing. 

In Fig. 171 is shown an example of metallic packing. The 
box or casing contains at its bottom, or upper end in the 
cut, a spiral spring, as shown. Next comes a brass ring, and 
next a series of babbitt or white metal rings carried in a casing 
or shell, as shown. These rings are conical on the outer cir- 
cumference, and fit to a corresponding form of the containing 


shell. Next below is a second brass ring which supports the 
shell above, the joint between the two being ground to a tight 
fit. This ring rests on a casing below, the joint between the 
two being spherical and ground to a fit. The latter cas- 
ing contains another spiral spring and then follows 
another series of rings, etc., similar to those above. The 
whole box contains, therefore, two similar sets of packing ele- 
ments, each consisting of a spiral spring, white metal rings con- 
taining shell, etc. Each of the white metal rings consists of two 
separate halves, the whole arranged so as to break joints from 
one ring to the next. It is readily seen that the action of the 
spring is to crowd the white metal packing rings into the conical 
shell and hence against the rod, thus keeping the joint tight be- 
tween the two. It is further seen that with this way of carrying 
the packing the latter is entirely unconstrained laterally, and 
may move in any way to accommodate itself to any slight ir- 
regularity in the rod without danger of disturbing the tight- 
ness of the joint. The two systems of packing, as a whole, are 
held up into place by an outer ring, secured to the cylinder head 
by stud bolts. The joint between the packing systems and the 
cylinder is made by a ring of copper wire, as shown, thus shut- 
ting off all leakage of steam from the packing space in this di- 
rection, while the various ground joints and packing ring sur- 
faces close it off in other directions. 

Among the various conditions which an ideal packing for 
piston and valve rods should fulfil those of chief importance may 
be stated as follows : 

(1) The packing should make a steam tight joint between 
rod and stuffing box, at the same time opposing the minimum 
frictional resistance to the motion of the former. 

(2) It must be durable even under the temperature of 
modern high pressure steam, and also easily removed or replaced 
with new when necessary. 

(3) The packing should be free to move about transverse- 
ly to a sufficient extent to follow the rod, even if it is slightly 
bent or out of line, at the same time maintaining the joints steam 
tight between the rod and the packing, and between the packing 
and the stuffing-box. 

Requirement (3) has given the greatest trouble and has led 
to many varieties of design intended to cover the point, one of 
which is illustrated as above in Fig. 171. No packing can be 


considered satisfactory for modern requirements which does not 
possess in good degree the qualities detailed above. 

[8] Reheaters. 

A reheater is a collection of pipes placed in a receiver or ex- 
haust passage from one cylinder to the next. Within these pipes 
high pressure steam is circulated, and around them the exhaust 
steam passes. The high pressure steam will, therefore, give up 
its heat to the cooler exhaust steam, and thus tend to dry or 
even to superheat it as it passes on into the next cylinder 
beyond. The office of the reheater is, therefore, to exercise a 
drying and heating action on the exhaust steam as it passes 
from one cylinder to another in a multiple expansion engine. 
Under most conditions this will exert a beneficial influence on 
the economy of the engine by decreasing the amount of cylinder 
condensation, and to such action may be referred the benefit 
which the reheater seems to give. See further Section 59. 

[9] Governors. 

In order to control the revolutions of the engine and to 
prevent violent increasing or racing when the propeller is par- 
tially lifted out of the water by the pitching of the ship, some 
form of governor is frequently fitted. The early types of mar- 
ine engine were usually governed by hand at the throttle, which 
was commonly of the butterfly variety (see Section 24 [i]), 
though occasionally automatic means of moving the valve were 
employed. With modern multiple expansion engines, however, 
it is impossible to satisfactorily control the revolutions by the 
throttle. With such engines the control must come from the 
slide valve gear, the links of which may be linked up more or 
less, as required when the propeller is uncovered, and linked out 
again as it is submerged. Where no automatic governor is 
fitted, this must be done by hand control of the reversing gear. 
The modern automatic governor is intended to take the place 
of this hand control. 

There are two modes of actuating marine governors. 

(1) By utilizing the varying pressure under the stern. 

(2) By utilizing a variation in the revolutions from the reg- 
ular speed. 

In the first type a pipe is run from the outside water at the 
stern to some form of pressure chamber near the engine, within 
which is a flexible diaphragm held in position by a spring or 


other equivalent means. The water being admitted to this pipe, 
the air within is compressed according to the head of water over 
the outer end. The apparatus is so adjusted that at normal draft 
the diaphragm is in equilibrium between the two forces, due to 
the water pressure on the one side and the spring on the 
other. A change in the depth of the water will cause a varia- 
tion in the pressure which will be transmitted through the air 
and thus destroy the equilibrium, throwing the diaphragm in 
one direction or the other. This may be made to actuate a 
steam valve and thus through an auxiliary steam piston control 
the reversing lever and through this the links. In former 
practice this type of gear was sometimes made sufficiently large 
to actuate the steam throttle directly, or sometimes a like valve 
in the exhaust pipe. 

The other type of governor is found in various forms. In 
many of them use is made of the centrifugal force of revolving 
balls or weights somewhat as in the ordinary stationary gover- 
nor. Through the force thus available a small valve is oper- 
ated, thus leading through a series of steps to the control of the 
reverse lever or other part which it is desired to operate. Again 
in other forms a revolving fan or propeller working in a box 
filled with liquid maintains the apparatus in a certain condition 
at a certain speed. With a sudden change of speed a corre- 
sponding change of resistance to the motion is met with, and this 
difference of force may be used to operate a small valve, and 
then, as before through the proper steps, the reversing lever is 
controlled. In another form a pump continually forces air into 
a chamber from which it escapes through a cock whose opening 
may be regulated at will. For a given size of outlet and speed 
of pump the pressure will rise until finally as much escapes as 
enters and the pressure remains constant. If the speed changes, 
however, the pressure will change correspondingly, and this 
difference of pressure will give a force which may be used as 
already explained. 

A still different type of gear operated by a change of speed 
but not driven by the use of a belt employs the forces due to 
inertia. As usually installed it consists of a weighted vertical 
rod pivoted at the top so that if unrestrained it could swing to 
and fro between a pair of stops. This weight with its point of 
suspension is then given a movement of reciprocation horizon- 
tally by attachment to any suitable part of the engine. If not 


prevented, it would therefore swing to and fro between the 
stops, due to the change in momentum imparted by the re- 
ciprocating motion. It is, however, held by a spring against one 
of the stops, and the tension is so adjusted that movement will 
not result until the engine exceeds its normal speed, when the 
inertia forces overcome the spring, and the weight moves away 
from the stop. This motion, by means of an attached lever, may 
operate an auxiliary valve, piston, etc., and thus control the 
links. This type of governor is sometimes used only as a safety 
gear to quickly stop the engine in case of a breakage of the shaft 
or other accident permitting violent racing. In such case the 
gear is sometimes so arranged that the movement of the weight 
away from the stop will cause it to engage as a clutch with an 
arm or lever connected with the reverse lever, and so adjusted 
that the motion given will just bring the links to the mid posi- 
tion. The instant the weight leaves the stop, therefore, the 
levers will be suddenly thrown over, the links middled and the 
engine stopped. 

All forms of marine governor are somewhat slow in control- 
ling the variations of speed. The ideal governor would antici- 
pate the motion and close down or open up just in advance of 
the rise in speed. Instead of this they act only after the stern has 
risen or fallen, or after the change of speed has become more 
or less pronounced. It is considered good practice, however, 
to fit some form of governor, at least as an emergency control, 
so as to prevent an excessive increase of revolutions from any 
cause whatever. For this purpose only those forms which de- 
pend on a change of speed are suitable, and such are commonly 
fitted in modern practice. 

[10] Counter Gear. 

The revolutions of the engine are automatically registered 
by a counter of the common type, and consisting of a series of 
discs with numbers from o to 9 on their circumferences. The 
motion for the counter is taken from any reciprocating piece 
which has a convenient location and a motion of small range. 
This is connected up to the counter by appropriate links, bell 
cranks, etc. The motion operates directly on the disc to the 
right and moves it along one notch or figure for each revolution. 
As each disc reaches o it engages with the one of next higher 
order on the left and throws it over, thus carrying the count 
continuously along the discs from the first to the last. 


In this way the revolutions are registered one by one, the 
total number for any period of time being found by taking the 
difference of the two readings for the beginning and end of this 
period. By this means the revolutions per minute, per hour or 
per day are readily found as desired. 

(n] Lagging. 

The cylinders, cylinder-heads, valve chests and covers are 
usually provided with some form of covering intended to prevent 
the loss of heat, and thus conduce to economy as well as render 
the engine less disagreeable to work near and about. Such 
covering is known as lagging and consists usually of either wood 
in strips or polished sheet brass or Russia iron. When of wood 
the strips are narrow, I to 2 inches in width, and often of alter- 
nate light and dark color to give a pleasing effect. They are 
usually matched together and secured by bands of brass or pol- 
ished iron, or by brass headed screws taking into foundation 
pieces held in place by countersunk tap bolts. 

[12] Lubrication and Oiling Gear. 

(i) Lubricants. For the lubrication of the various rubbing 
surfaces and turning joints, except within the cylinder, olive oil, 
castor oil, and the lubricating grades of mineral oil are used. 
Olive oil when pure is a most excellent lubricant, especially for 
machinery of moderate or light weight, but it is liable to adul- 
teration by peanut oil or other oils of an inferior quality. For 
the lubrication of the internal surfaces nothing but the best min- 
eral oil must be used. (See Sec. 40.) The grade commonly 
employed is known as cylinder oil, and is heavy and viscid at 
ordinary temperatures, becoming quite fluid, however, at the 
usual temperatures of the steam. It is usually fed in by means 
of a sight feed lubricator, as described in (9). 

In addition to the liquid oils, various lubricating greases are 
used, often in combination with a certain proportion of graphite 
(black lead.) Graphite alone or in combination with oil 
is also used, and its lubricating qualities are of the highest 
order. It seems to possess the property, especially with cast 
iron, of filling the pores of the iron, and of thus forming a kind 
of graphite-metal skin on the surface, with a very small co- 
efficient of friction. 

The place where the lubricant should be supplied to the 
bearing is a subject which has attracted considerable attention 


in the past few years. It seems now to be very well established 
that the following principles should govern : 

(a) The oil should always, where possible, be led into the 
bearing at a point which is under the smallest pressure. 

(b) The continuity of the oil film, where it is under the 
greatest pressure, should not be interrupted by oil channels or 

(c) The oil should be prevented, as far as possible, from 
escaping at those points which are under the greatest pressure. 

For journals such as main pillow-blocks, etc., these princi- 
ples are very commonly violated, and in fact it can hardly be 
said that practice has as yet come to act upon them, though their 
correctness seems to have been well demonstrated. According 
to these principles the oil for the main pillow block bearings 
should be introduced near the division between the upper and 
lower brasses, and the oil scores or grooves in the metal of the 
bearing at the top and bottom should be omitted. Similarly for 
the crank-pin and other cylindrical journals, the oil should be 
admitted at those points where there is the least pressure, and 
at the points where the pressure is greatest the bearing surface 
should be smooth and not interrupted by grooves, or scores, or 
oil channels of any kind whatever. 

(2) Amount of Lubricant Required.. As to the amount of oil 
to be used, practice differs widely, but from 5 to 8 Ib. of oil per 
ton of coal may be taken as a fair allowance, or say from 5 to 8 
Ib. per looo I. H. P. per hour. In small sizes, for engines of the 
torpedo boat type, etc., the consumption will go up to consider- 
ably larger figures. Of this total amount some 5 to 10 per cent, 
may be required for internal lubrication, and the remainder for 
the various joints, bearings, etc. 

In the opinion of many good engineers, after a ship has 
been at sea for a few days and the machinery has settled into a 
steady running condition, the amount of internal lubricant may 
be gradually decreased to perhaps one-half the amount first 

With regard to the frequency of lubrication no definite rules 
can be given. The ideal system is, of course, as nearly con- 
tinuous as possible. Where, however, the continuous system is 
not in use and intermittent oiling must be depended on, the vari- 
ous joints and stuffing boxes will require attention and a fresh 
supply of lubricant at intervals of from perhaps twenty minutes 


to one hour. For the various pin and turning joints, main 
guides, etc., the lubricant is usually supplied by oil cups or cans 
of character suited to the particular use for which intended, 
while the piston and valve rods are lubricated by means of a 
swab charged with cylinder oil or special grease. The main 
guides are also in some cases lubricated by the swab rather than 
the can. 

There is also a growing tendency in engines of the high 
speed type as used on torpedo boats, etc., and where the steam 
is always more or less moist, to depend on water lubrication, 
and to avoid, so far as possible, the use of internal lubricant. 
This end may be furthered by careful workmanship in the fitting 
up of valves and pistons, and by driving the engine by belt or 
otherwise in the erecting shop with the surfaces charged with 
graphite. The minute pores of the metal are thus filled with the 
graphite, and rubbing surfaces are developed which run very 
well without further lubrication. 

(3) Adjustment of Bearings. For a bearing in good adjust- 
ment the clearance or distance between the journal and bearing 
surface is proportioned to the size of the journal, and may be 
made about .002 or 1-5 of i per cent, of the diameter. With a 
lubricant of proper consistency and a load per square inch not 
too great, say not over 400 to 500 Ib. per square inch of project- 
ed area, the film of oil will retain its place, and insure the proper 
lubrication of the bearing. If too thin a lubricant is used the 
bearing may heat and pound simply because the journal is not 
supported by the film of oil. The proper consistency of the oil 
as influenced by its natural viscosity and by the temperature of 
the bearing may, therefore, determine to a considerable extent 
the smooth running or pounding of the various joints and jour- 


We will now describe the more important devices for sup- 
plying oil and grease to the bearings, or to the points where re- 

(4) Wick Cup. The plain wick cup consists of a receptacle, 
usually of cast brass, fitted with a cover, and placed at a conven- 
ient point for the delivery of the oil to the bearing. It is also 
frequently formed as a part of the bearing cap, as in Fig. 147. 



A tube, as shown, enters through the bottom of the oil reservoir 
and rises within to a point above the level at which it is expected 
to carry the oil. This tube leads downward to the duct which 
carries the oil to the point of delivery to the bearing. The 
"wick" consists of a few threads of cotton wicking, one end of 
which is wrapped with a bit of wire, which then serves as a 
handle for pushing it down the tube or for pulling it out. In op- 
eration, one end of the wick is pushed down the tube and the 
other end dipped in the oil. Through the action of capillary at- 
traction the oil rises in the wick on the outside, and then by a 
combination of capillary and siphon action descends and drips 
down the tube to the bearing. The end of the wick within the 
tube should be pushed down below the level of the bottom of the 
cup so that this shall form the longer leg of the siphon. 

The size of the wick should be adjusted according to the 
amount of oil which it is desired to feed and to the quality of the 
oil as well. This adjustment of size is most easily effected by 
varying the number of strands of cotton in the wick. The 

Fig. 172. Wiper. 

Fig .173. Oil Cup with Adjustable Feed. 

amount fed may also be varied to some extent by regulating the 
distance to which the wick is pushed down the inner tube. 

For a sudden flush of oil the cups may be filled until they 
overflow into the inner tube, by which means the bearing may 
be flooded if desired. 

(5) Wiper. A wiper is shown in Fig. 172. It consists of 



an oil cup with a central blade or plate, A, extending above the 
edge, and attached to one of the moving parts of the engine. At 
a convenient point is placed a strip of fibrous material on to 
which the oil is fed from the source of supply. The strip and 
wiper are so adjusted that the latter in its motion to and fro 
wipes or scrapes along the lower surface of the former, and thus 
as soon as the strip is saturated with oil the wiper takes off a 
drop or more which then runs down into. the cup and thence to 
the surfaces to be lubricated. Naturally this mode of lubrica- 
tion is more especially suited to parts having a horizontal mo- 

Fig. 174. 

Marine Engineering 

Oil Cup with Screw Cover. 

Marine Engineering 

Fig. 175. Oil Cup with Sight Feed. 

(6) Plain Oil Cup With Adjustable Feed. This consists of a 
simple cup, as shown in Fig. 173, mounted where convenient 
and connected by pipe or duct to the bearing to be supplied. 
The rate of feed is regulated by a needle or conical valve, which 
controls the size of opening through the discharge passage in 
the base. A cover is usually fitted to prevent spilling or the ad- 
mission of impurities. Where such a form of cup is used to ad- 
mit oil to a steam chest or other chamber under pressure, a 
strong screw cover is necessary, as shown in Fig. 174. To fill 
the cup in such case the valve is closed, the cover unscrewed, and 


the cup filled. The cover is then replaced and the valve opened 
according to the rate of feed desired. ^ Another variety of cup 
used for this purpose has a body of cylindrical or globular form 
terminating at top and bottom in a neck or tube, each of the 
latter being closed by a valve. The lower neck joins the tube 
leading to the bearing as in other cups, while to the upper is 
fixed a shallow open cup. The chamber between the two 
valves is filled by closing the lower valve, opening the upper 
and pouring into the shallow cup which serves thus simply as a 
funnel. When filled, the upper valve is closed and the lower one 
opened according to the rate of feed desired. 

(7) Sight Feed Oil Cup. As shown in Fig. 175, this is es- 
sentially a plain cup with the addition of the "sight-feed" attach- 
ment or feature. When in adjustment, the flow of oil is regu- 
lated by the conical valve to a drop at a time at such interval 
as may be desired. The space below the outlet of the cup is 
cut away so as to show the drop as it falls into the mouth of the 
feeding tube. Very frequently a glass tube is fitted inside the 
brass framework, thus closing in the oil completely, but allowing 
the drop to be seen as it falls. 

For a sudden flush of oil it is only necessary to open up the 
conical valve sufficient to let the oil descend in a stream and 
flood the bearing. 

(8) Compression Grease Cup. In addition to oil, various 
forms of hard grease in cakes, or balls, or in bulk, are some- 
times used for lubricating purposes. For feeding such lubri- 
cating material to bearings two means are made use of. 

(i) If the bearing tends to become heated the heat de- 
veloped will soften the grease and allow it to run to the spot 
where it is needed. (2) Compression cups are used containing 
a piston or plunger on top of the grease and acted on by a 
spring under control by a screw operated by hand. (See Fig. 
176.) The grease is thus forced either automatically or by hand 
through the feeding tube and to the bearing. The spring ar- 
rangement may be made adjustable so as to force the grease 
more or less rapidly, according to its degree of hardness, and to 
the rate of feed desired. 

(9) Lubricator. For the introduction of cylinder oil into the 
valve chests or steam pipes, an apparatus known as a sight-feed 
lubricator is very commonly employed. Such devices have been 
made in great variety of form, but the description of one will be 


sufficient to show the principle upon which they operate. The 
lubricator, shown in Fig. 177, consists of a main chamber, A, 
with connections for attachment to the steam pipe or throttle 
valve casing at B, and for the attachment of a length of vertical 
pipe, P, leading in to the main steam pipe at C. D and E are 
two fittings for a short length of glass tube, as shown. From 
the top of the chamber a passage or pipe leads down to the lower 
fitting, E, while from D the passage leads into the steam pipe 

Marine Engineering 

Fig. 176. Compression Grease Cup. 

Fig. 177. Automatic Lubricator. 

through the connection at B. The lower part of the chamber is 
connected to the vertical pipe leading up to C. There are also 
two connections, F, G, with glass gauge between to show the 
level of the oil and water within the chamber. The operation 
of the lubricator depends on the difference in density between 
the oil and water. The lubricator is first filled with oil through 
the plug H. The steam is then admitted to the pipe, P, where 
it will slowly condense and collect at the base of the chamber, 
in the pipe between D and E, and in the pipe P, thus furnishing 
a head of water acting on the oil and forcing it upward. As 


soon as the head is sufficient, oil will be forced a drop at a time 
as regulated by the valve V, out through the passage leading 
from the top of the chamber down to E and up through the 
water in the tube, DE, and so on to the steam pipe, where it is 
caught by the flow of steam and carried to the valve chest and 
cylinder. The passage of the oil drop upward is plainly seen, 
and thus the operation of the lubricator is under ready observa- 
tion. Such lubricator may be placed on the steam pipe or throt- 
tle valve chamber, or at any convenient point where the oil will 
be carried by the inflowing steam to the points where needed. 

(10) Oil Pump. A simple arrangement for forcing cylinder 
oil into a steam chest is sometimes used when it is not conven- 
ient to fit a lubricator. This consists, as shown in Fig. 178, of 


Fig. 179. Oil Pump. 

an oil pump operated by hand. The chamber being filled with 
oil the delivery valve is opened and the oil forced in as may be 
desired, through a connection attached to the pump delivery, as 

(n) Modern Systems of Oil Distribution. In the preceding 
section we have described the principal devices used for supply- 
ing oil to a bearing, or to a steam pipe or chest. We will now 
describe briefly the general oiling system for the engine as a 
whole, involving such combination of these devices as may be 
found most desirable. 

The leading features of the modern system consists in the 
provision of a small number of distributing centers from which 
oil is taken by piping as directly as possible to the various places 


where needed, each place having its own independent pipe and 
set of connections. Following is a brief description of such a 
system of oiling gear, and will serve to illustrate the methods 
now in use in good practice. 

A light cast brass box is provided for each cylinder placed 
at a point higher than any joint or bearing to be reached by the 
oil, and having a capacity sufficient to last several hours without 
refilling. These oil boxes are provided with sight feed cups with 
protected glass tubes from which pipes lead to wipers on the 
moving parts, or to tubes in the bearings and guides. Union 
joints are fitted where necessary, so that the oil pipes may be 
quickly taken down and cleaned. With few exceptions the oil 
for the various moving parts of the cylinder is supplied from 
this box. 

The main crank pin is oiled by means of a pipe and cup 
carried on the cross-head and taking oil from a drip supplied 
from the oil box as described. The pipe runs down the side of 
the rod, or frequently inside if the rod is hollow, and connects 
with the oil duct leading through to the pin. The cross-head 
guides are provided with oil through pipes connected with holes 
at about the middle of each forward and backing guide. The 
main pillow-blocks are oiled by one or more wick cups deliver- 
ing the oil at the points desired. 

A cup for tallow or grease is also usually provided, and like- 
wise sometimes a hole through which the hand may be passed 
to feel of the shaft as may be desired. The presence of this 
hole, however, is not in accord with the principles given above 
in (i), and the practice cannot be recommended. If anything of 
the kind is to be fitted it is better to carry the hole simply 
through the cap, thus leaving the brass continuous. The latter 
may then be felt as desired, and a tendency to heat may be thus 
observed. The excentric straps are fed from long narrow oil 
cups, receiving their oil through the drip pipes from the reser- 
voir. The length of the cup is made such that some part of it 
is always under the drip in any position of the excentric. and it 
will, therefore, always receive its supply. The various other 
parts of the gear are similarly supplied with oil either from a 
drip or a wiper as may be more convenient. 

The chief advantages of such a system consist in the cer- 
tainty and regularity of operation which may be assured with the 
minimum of time and attention on the part of the oiler. 


Sec. 25. PIPING. 
[i] Systems and Materials. 

The principal systems of piping are as follows : 

(1) The main steam piping from boilers to engine. 

(2) The auxiliary steam piping from the auxiliary boiler 
or from one or more boilers specially selected, to the various 
auxiliaries which are to be operated by steam. 

(3) The main exhaust piping from cylinder to cylinder and 
from the L. P. cylinder to the condenser. 

(4) The auxiliary exhaust system, providing each auxil- 
iary with its exhaust, either to the condenser or overboard, as 

(5) The feed system, main and auxiliary for returning the 
condensed steam to the boilers. 

(6) The condensing system for bringing the condensing 
water to the condenser and for carrying it from the condenser 
to the sea again. 

(7) The drainage and bilge pump delivery systems for lead- 
ing water to the bilge pumps and from the pumps to the sea. 

(8) The fire system for leading water to the fire pumps and 
from the pumps to the fire plugs. 

(9) The sanitary system for delivering water to the W.Cs. 
and heads. 

(10) The steam heating system. 

We may otherwise classify all pipes under three heads : 
steam pipes, exhaust pipes, and water pipes, while of the latter we 
have a further division into those carrying water to a pump or 
induction pipes, and those carrying water from a pump or educ- 
tion pipes. 

For steam pipes the materials in present use are copper, 
wrought-iron, and steel. Copper pipes in small or moderate 
sizes may be made of seamless or solid drawn tubing ; in large 
sizes they are made of sheet copper with brazed joints. 
Wrought iron pipes are lap welded, while steel pipes are also lap 
welded and are sometimes further fitted with a riveted longi- 
tudinal strap covering the line of the weld. Seamless or solid 
drawn steel pipes have also been made to some extent. 

For the various junctions, elbows, bends, tees, etc., steel 
or malleable iron castings are used with steel pipe, while with 
copper pipe sheet copper is used for these parts, bent and formed 


up by hammering into shape, and secured at the joints by 
brazing 1 . 

The advantages of copper are its great ductility, freedom 
from corrosion, and the readiness with which it may be used to 
make pieces of an irregular form, such as the elbows, junctions, 
etc., referred to above. Its disadvantages are, greater cost, low 
tensile strength, the possibility of damage to the quality of the 
material in the process of pipe manufacture, and the possibility 
of a loss of ductility in service by repeated strains due to the 
expansions and contractions which result from changes in tem- 

The metal close about a flanged joint seems especially 
liable to lose its strength in this way. This is probably due to 
the concentration of the strains due to expansions and con- 
tractions in the vicinity of a rigid connection such as a flanged 
joint, and to the development in this way of a line of weakness 
running around the pipe. 

To render copper pipe more secure under high pressure it 
has been wound with copper or steel wire, or reinforced by 
wrought-iron or steel bands. Such bands may be from ^ to 2 
inches wide by y& to J4 inch thick, and spaced with intervals of 
from 6 to 10 inches. These methods, especially the latter, have 
proven quite successful in strengthening copper piping for mod- 
ern advancing pressures. 

The chief advantage of wrought iron and steel pipes are, 
less cost, greater tensile strength, less liability of the mater- 
ial to damage in quality in the processes of manufacture, and 
less liability to lose strength or ductility in service. Their 
disadvantages . are, greater liability to corrosion, and greater 
weight of cast junctions and fittings than for copper. 

Welded pipe is made from rolled strips the edges of which 
are machine beveled for a lap joint. The requirements of manu- 
facture are such that for all except the largest sizes the 
thickness is in excess of that needed for strength alone, at least 
with the pressures at present in use, so that when such material 
is employed there is always an excess of strength. 1 

With wrought-iron pipes the welded joint is trusted without 
reinforcement. With mild steel a covering strip or butt strap 
is sometimes riveted on, though with the later improvements in 
the welding quality of such material, experience shows that these 
joints are quite as reliable as those of iron. Flanges for 



wrought-iron or steel pipes may be welded on, but more com- 
monly they are riveted or screwed to the pipes. In the latter 
cases the flanges are caulked on both sides to make a steam 
tight joint. 

In small sizes and in a relatively cheaper grade of practice, 
ordinary commercial steam piping is used, fitted up with the 
usual fittings and screwed joints. 

For exhaust piping the same general character of pipe is 
used as for steam, with such differences as the decreased strength 
necessary may indicate. 

For water piping steel, iron and copper, and in cheaper 
practice the ordinary commercial pipe are all used. Steel and 
iron are usually considered less suitable for water than for 
steam piping on account of the greater danger from corrosion. 
This is especially true for feed piping, and in case such material 
is used for this purpose it is considered good practice to care- 
fully galvanize the pipe both inside and out. It is likewise good 
practice to tin all copper piping which is under the floor plates 
or in the bilge, but this is rather to protect the ship than the 
pipe, the former being in danger of attack by electro chemical 
action (see Sec. 40) in case the copper and the metal of the 
ship obtain connection through a medium such as bilge water. 
In connection with piping see also that heading under Section 19. 

[2] Expansion Joint. 

The expansion and contraction of a length of piping under 
a change of temperature require some kind of joint or connec- 

Fig. 179. Expansion Joint. 

tion which will allow of the change of length without buckling or 
straining the pipe. This is usually provided by an expansion joint, 
as shown in Fig. 179. This consists of a recessed portion on 
one part of the joint into which the other fits, as shown. The 


space left between the two thus forms a stuffing-box into which 
packing is compressed by means of the gland. The two parts 
of the joint are thus free to slip a little way, one relative to the 
other, while the joint is kept tight by means of the stuffing box 
and gland in the usual manner. It is readily seen that the steam 
pressure within the pipe, especially if it contains a bend or el- 
bow, will tend to force the two portions of the joint apart, and 
thus open the pipe at the joint. To guard against this, safety 
stays or ties should be fitted. In the figure one of three such 
stays is shown by the bolt at the top. Care must be taken in 
adjusting such stays that sufficient freedom is left for the expan- 
sion and contraction which the joint is intended to provide for. 

In special and more complicated forms, known as balanced 
or equilibrium expansion joints, these forces are more or less 
completely balanced within the joint itself. 

[3] Globe Angle and Straightway Valves. 
For controlling the flow of a liquid, vapor or gas through a 
line of piping, various forms of valve are used. Chief among 
these are the Globe, Angle and Straightway or Gate types, as de- 
scribed in Section 24 [i], [2], and to which reference may be 






The office of the circulating pump is to draw the condensing 
water from overboard, force it through the condenser tubes, as 
explained in Section 29, and thence overboard through the con- 
denser discharge pipe. The principal resistance to be 

n/w Engineering 

Fig. 180. Centrifxigal Pump. 

come by the pump is the resistance to the llo\v of the water 
through the tubes, and this is but slight measured in pounds per 
square inch or in feet of head. On the other hand, the quantity 
of water to be handled is large, and hence the requirement is 
for a type of pump which shall be able to handle large quantities 
of water against a small head or resistance. These require- 
ments are very perfectly fulfilled by the centrifugal pump, as 
shown in Fig. 180. The moving part consists of a number of 



vanes or arms attached to a shaft and forming what is called 
the runner. This revolves within a casing furnished with inflow 
and outflow passages, as shown in the figure. The pump being 
primed or filled with water and started, the rotary motion gives 
rise to a centrifugal force, in obedience to which the water moves 
outward toward the tips of the blades, where it escapes through 
the outflow passage into the discharge pipe. There is a corre- 
sponding defect of pressure about the hub of the runner and a 
resulting inflow of water from the sea to take the place of that 
which leaves at the outflow. The operation thus becomes con- 

Fig. 181. Outboard Discharge Valve. 

tinuous and results in a steady flow of water from the pump 
through the condenser tubes and back to the sea through the 
condenser discharge pipe and outboard delivery. In order to 
prevent the water from "short circuiting" or slipping back from 
the discharge space about the tips of the blades to the inflow 
space about the hub, a running fit is provided between cor- 
responding faces on the runner and casing as indicated in the 

The discharge or outboard delivery valve is usually a plain 
type of angle stop valve, as illustrated in Fig. 181. Its office is 


simply to allow when open the discharge of the water from the 
condenser, and prevent when closed the inflow of the sea to the 


The purpose of the condenser is to provide for converting 
the exhaust steam back into water. Condensers are of two 
types jet and surface. 

The jet condenser consists simply of a chamber of rec- 
tangular or cylindrical form in which the steam and the conden- 
sing water are mingled together, the steam giving up its heat to 
the relatively cool water, and thus being reduced to the liquid 
state again. The water is usually led into the top of the 
chamber and allowed to fall upon a plate pierced with a large 
number of small holes, and known as the scattering plate. This 
divides the water into small streams or jets and enables it to mix 
intimately with the steam which enters just below the plate. The 
condensed steam and condensing water then fall together to the 
bottom of the condenser. From here the water is removed 
by the air-pump which delivers it overboard, the feed-pump in 
the meantime taking enough for the boiler feed and returning it 
to the boilers. 

The usual type of surface condenser consists of a chamber 
commonly of cylindrical form if separate from the main engine 
(see Fig. 182), or rectangular if forming a part of the engine 
columns. (See Fig. 100.) This chamber contains, as shown, 
a large number of small brass tubes running between the inner 
walls of the heads, which are double, thus providing a connec- 
tion between the ends of the various tubes. The condensing 
water is driven by the circulating pump through the tubes, the 
usual run being, as shown in the figure. The water enters at 
the lower left hand end and fills the lower half of the head, be- 
ing prevented from filling the whole head by a partition half way 
up, as shown. It thus finds its way into the lower half of the 
tubes and flows through them to the right hand head. It then 
rises into the upper half of the tubes, flows back to the left and 
out at the opening in the upper part of the left hand head. The 
water thus traverses twice the length of the condenser forward 
and back, and from the bottom upward. The steam, on the other 
hand, enters at the top into the body of the chamber, and thus 
around the outside of the tubes. The steam and the condensing 
water are thus kept separate, and the steam is condensed simply 



by the surface action of the tubes. The steam thus condensed to 
water falls to the bottom of the condenser, whence with some air 
and vapor it is removed by the air-pump and delivered to the hot- 
well, whence it is taken by the feed-pump and sent back to the 
boilers. Baffle or diaphragm plates, as shown, are often fitted in 
the condenser to prevent the steam from rushing directly 
through from the inflow on top to the air pump passage on the 
bottom. The steam is thus forced to fill the condenser as com- 
pletely as possible, and thus the condensing surface is more uni- 
formly brought into action. In order to facilitate the rush of 
steam downward into the body of the tubes, thus bringing more 


- Fig. 182. Surface Condenser, Longitudinal Section. 

quickly into action those in the lower part of the condenser, a 
few rows are often omitted in the upper part, thus forming 
branching passages leading from the top downward, as shown in 
Fig. 183. In order to support the heads against the pressure 
from without, a certain number of longitudinal braces or struts 
are necessary, as shown in the figures. 

Condenser shells are made of cast-brass, cast-iron, or sheet 
brass or steel. When of rectangular form the sides are cast 
with the necessary webs to give them strength to stand the pres- 
sure from without. When of cylindrical form the necessary 
strength can be given bv a suitable thickness of metal rein- 


forced if necessary by ribs running around the shell. When rela- 
tively thin sheet metal is employed, as in torpedo-boat practice 
and the like, it is customary to fit one or more angle-iron or 
Tee-iron stiffeners running around the shell in order to provide 
the necessary strength. 

Condenser tubes are of thin sheet brass, usually > to y^ 
inch outside diameter. In order to make a water-tight joint be- 
tween the condenser tubes and the inner heads or tube plates, 
and at the same time to avoid a rigid constraint of the tube, a 
great variety of condenser tube packings have been employed. 
The most common type of packing in present practice is shown 
in Fig- .184. The tube plate is counterbored, as shown, and 
threaded for a ferrule with a tapering outer end. The hole in 
the outer end of the ferrule is thus of about the same size as 


3fartnt Engineering 

Fig. 183. Surface Condenser, Cross Section. 

that inside the tube, and hence smaller than the outside of the 
tube. Between the other end of the ferrule and the bottom of 
the counterbore is usually a ring of rubber or a few turns of some 
other elastic or fibrous material as packing. Screwing clown on 
the ferrule compresses the packing, and thus makes the joint, 
while at the same time the tube is free to expand and contract 
to a slight extent. The outer ends of the ferrules, however, pre- 
vent the tube from crawling to such an extent as to free either 
end, a result liable to occur without some method of prevention. 

Sec. 28. AIR PUMPS. 

It is. the purpose of the air-pump to remove from the con- 
denser the water and such small quantities of air as may enter 
by leakage or with the steam, and which would ultimately de- 

2 4 6 


stroy the vacuum if riot removed. As often stated, it is the 
office of the air pump to maintain the vacuum formed by the 
condensation of the steam. 

The usual type of air-pump is shown in Fig. 185. A is the 
piston or bucket moving in the barrel B, and carrying bucket 
valves, D, opening upward, as shown. The foot valves at C also 
open upward and admit the contents of the condenser from 
below. Beginning with the piston in the position shown the 
operation is as follows : 

As the piston rises, the air and vapor between its lower 
face and the foot valves become rarified with a resultant decrease 

Frg. 184. Ferrule and Tube Packing, Surface Condenser. 

of pressure. Soon a point is reached where the pressure in the 
condenser is decidedly greater than in the space above the foot- 
valves, and in answer to this difference of pressure the valves 
open and admit air, vapor and water from the condenser. This* 
operation terminates with the piston at the top of the stroke. On 
the return stroke the foot valves close and the contents of the 
barrel are forced through the bucket valves at D to the space 
above. On the next stroke the contents are lifted and forced 
out through the delivery or head valves at E, w r here the air and 
vapor escape, and the water flows to the hot-well, whence it 
is sent by the feed-pump back to the boiler. 



Fig. 185. Vertical Attached Air Pump. 

2 4 8 


It is evident that the pressure in the condenser cannot be 
reduced below that necessary to raise the foot-valves, so that 
for this reason, as well as for their more ready response to varia- 
tion of pressure, they should be made as light as is consistent 
with a proper performance of their duty. As shown in Fig. 186, 
such valves in modern practice are usually made of light sheet 
metal discs controlled by spiral springs. In older practice vul- 
canized rubber was quite commonly employed. It is also evi- 
dent that the foot valveb will respond the more quickly, the more 
rapidly the pressure above them decreases as the piston begins to 
rise, and hence the less the clearance space between the valves 
and the piston when in its lowest position. 

The air-pump has this peculiarity in its action, that the 
load per stroke and hence the resistance often dUreases with an 
increase of speed, and increases with a decrease of speed. Hence 
when the pump is operated by an independent engine, it may 

Fig. 186. Air Pump Valve, Guard and Spring. 

be liable, unless carefully designed, to race or run away to ex- 
cessively high speeds with an increase of pressure in the steam 
cylinder, or to slow down and stop with a corresponding de- 

It is apparent also that a certain time will be needed for 
the foot valves to open, and for the air, water and vapor to flow 
through. Hence, with excessive speeds, the valves may not have 
time to open between strokes, the vacuum will become poorer, 
'and the condenser may become flooded with water. It thus fol- 
lows that with too high, as well as with too low a speed, the 
vacuum will be poor, and the operation of the pump unsatis- 

The air-pump may be driven either by an independent en- 
gine, or by attachment to the main engine. The attached air- 



pump is usually operated by air pump levers, which derive their 
motion in most cases from the L. P. cross-head, as shown in Fig. 
<*<;. The stroke of the pump is thus reduced to usually less 
than one-half that of the main engine. One of the advantages 
of the attached air-pump is that the number of strokes per min- 
ute is necessarily the same as for the main engine, and it can 
neither race nor slow clown on account of variations in its own 
resistance. A further advantage is that the power required for 
its operation is obtained more economically than when operated 
by a separate engine. The chief disadvantage of the attached 
air-pump is that with the modern increase of revolutions, the 
speed may be too great "for the best results from the pump as 

Fig. 187. Air Pump for High Speeds. 

usually designed. This, together with the advantage of having 
the condition of the condenser under control independent of the 
main engine furnish the chief reasons for the use of the inde- 
pendent air-pump. When thus fitted as an independent auxil- 
iary the number of double strokes per minute usually varies 
from 15 or 20 to 30 or more, while the revolutions of the main 
engine may be from 100 to 200 or more. 

For special cases where the revolutions are very high, as in 
torpedo boat practice, but where for simplicity or for the saving 
of space it is desirable to use an attached air-pump, the Bailey 
type of pump is employed. In this pump, as shown in Fig. 187, 



the water flows by gravity into the barrel through ports alter- 
nately opened and closed by the piston itself, which thus serves 
as its own valve. The air and vapor naturally expand and enter 
with the water, and the whole contents are forced out of the 
end of the barrel through delivery valves similar to those in the 
pump of usual type. In some cases the delivery valves are car- 

Fig. 188. Vertical Independent Air Pump. 

ried on movable heads, which thus become valves in themselves, 
and available to relieve the barrel in case the smaller valves give 
insufficient opening. In some cases, as shown in Fig. 187, the 
smaller valves are omitted and the entire head serves as the de- 
livery valve. With this design air-pumps may be successfully 
operated at speeds of 400 or 500 revolutions and higher. 


When operated independently the air pump is very com- 
monly made double, and operated by one form or another of 
special steam valve gear. A typical form of independent air 
pump is shown in Fig. 188. The general operation of the valve 
gear is as follows : The beam which positively connects the main 
piston rods of the pumps operates from a point near its center 
and by means of rod and bell cranks, the slide valve of the hori- 
zontal cylinder which lies between the main steam cylinders, as 
shown. The piston of this horizontal cylinder is really the driv- 
ing engine of the main steam valves, a function which it per- 
forms by means of a system of internal levers. See further Sec- 
tion 33 regarding the operation of such types of pump valve 
gear. The adjustable collars on the valve stem of the 
"valve driving engine" afford a means for regulating 
for full stroke at any speed, while suitable cushion valves give a 
further control over the action during the stroke, in regulating 
the distribution of the work and preventing the slamming of 
the foot valves. 


Comparing the feed and circulating pumps we find that 
the former has to handle a very much smaller quantity of water 
usually from 1-25 to 1-40 the amount but against a very much 
higher pressure, viz., that in the boilers. In consequence an 
entirely different type of pump is required. 

The feed-pump may be attached to the main engine, or run 
as an independent auxiliary. When attached to the engine it 
is usually of the type known as the plunger pump and shown in 
Fig. 189. The moving part consists simply of the plunger, AB, 
working in the stuffing box, KL, and operated usually from the 
air-pump levers. There are two valves or sets of valves, in- 
flow and outflow, as shown at F and E. The level of the pump 
is usually below that of the hot-well so that the water stands 
ready to enter through the inflow valves as the plunger rises and 
makes room for it. This is aided by the partial vacuum formed 
within the barrel as the plunger rises. On the other hand, as the 
plunger descends on the next stroke the inflow valve closes and 
the water flows out through the outflow valve in order to make 
room for the descending plunger. It is thus seen that the pump 
is single acting ; that is, that it delivers but once in two strokes, 
and that the amount delivered is measured by the volume of the 
plunger displacement. 



This in turn equals the cross-sectional area of the plunger 
multiplied by the length of the stroke. The actual delivery per 
stroke will be somewhat less than this due to leakage, and to 
failure of the barrel to completely fill on the up stroke. 

The stuffing box, K L, is of course accessible and adjustable 
from the outside, and with proper design the inflow and outflow 
valves may be examined by the simple removal of a bonnet. 
The strong points of this pump are its simplicity, and the ready 
accessibility for examination and adjustment, of all parts on 
which the operation of the pump may depend. 

Fig. 189. Plunger Feed Tump. 

In fitting up the attached feed pump it is necessary to pro- 
vide it with some form of safety or relief valve, else should the 
discharge or check valve jam or fail to operate, the feed pipe 
or pump or some part of its operating gear would become 
broken. Such a relief valve is usually a simple form of spring 
loaded safety valve, and similar to the engine relief valves re- 
ferred to in Section 24. In order that the valve may be ef- 
fective in relieving the pump and entire line of pipe it should be 
placed on the pump chamber or in any event not beyond the 
pump discharge valve. 


Where the feed pump is operated as an independent auxil- 
iary it is usually of the direct acting or positive motion type, as 
described in Section 33, and to which reference for details may 
be made. For feed-pump purposes the area of the steam piston 
is made from 2 to 3 times the area of the water piston or plunger 
in order to give on the steam side a pronounced excess of total 
pressure over the resistance on the water side. This will enable 
the pump to overcome the resistance to the flow of the water 
through the feed-pipe and check valves, and thus to force water 
into the same boiler from which it draws its steam, or even into 
a boiler with a pressure somewhat higher than that from which 
its steam is drawn. For the general purposes of a feed pump 
the vertical or admiralty style, as illustrated in Section 33, has 
come to be very generally used. Its chief advantages over the 
horizontal type are two in number. 

(1) It occupies less floor space and may be conveniently 
put up on a bulkhead or elsewhere, in such manner as to occupy 
but little space otherwise available. 

(2) The valves in the water end, as shown in the figure, 
are more conveniently arranged for examination by the removal 
of a bonnet than with the horizontal type of pump. 

These considerations and especially the latter, have made 
this general type of feed pump the standard in modern marine 
engineering practice. 

In addition to feed-pumps of the plunger and piston types, 
an injector is often fitted as an auxiliary means of feeding the 
boilers. There are many different varieties of injector, but a 
description of one will suffice to illustrate the principles in- 
volved. Referring to the diagram. Fig. HJO. S. is a nozzle con- 
nected with the upper pipe, I>, leading steam from the boiler. 
When steam is turned on by means of the handle, K, and at- 
tached valve-stem and valve, it escapes in a jet which enters the. 
slightly tapered passage VC. The air in the space around and 
between these two orifices is caught and drawn along with the 
jet, thus causing a reduction of pressure at this point. This 
space is connected through the lower pipe, B, to the water reser- 
voir, and when the loss of pressure is sufficient, the water rises 
the same as in the case of a pump. The water and steam are 
thus brought into contact, and pass on together into the com- 
bining and delivery tube CD. The steam is here condensed and 
the resultant jet of water attains a very high velocity. A little 


further on, when this is reduced to the relatively low velocity of 
the water in the feed pipe, the pressure developed is sufficient to 
overcome the boiler pressure, to open the check valve, and to 
force the water into the boiler. 

It may aid the understanding of this seemingly puzzling re- 
sult if we remember that it is the energy of the steam which is 
the real motive power. This is transformed largely into motion 
in the combined jet, and this, when arrested, gives the pressure, 
as stated before. In the steam pump a steam piston is provided 
much larger than the water plunger in order to give a force suffi- 
cient to overcome the resistance of the feed pipe and at the 
check valve. So in the injector, in a somewhat similar manner, 

Fig. 190. Injector. 

the energy of the steam in a relatively large pipe is concen- 
trated on a small jet of water, giving it the high velocity and later 
the pressure, as described. 

In passing it may be noted that as a boiler feeder a good 
injector has practically a perfect efficiency, all heat used being 
carried back again into the boiler, except the small amount lost 
by radiation from the instrument and connecting pipes. 

An injector of the type shown in the figure is known as an 
automatic injector. This signifies that once .the injector is ad- 
justed and working, should the jet of water become broken by a 
jar or other accidental circumstances, it will restart itself without 
further adjustment. The capacity and working range of an in- 


jector are decreased as the lift is higher and as the water is 
warmer. With cold water and a moderate lift, say not exceed- 
ing 5 to 8 feet, a good automatic injector will start up with 25 or 
30 Ib. steam pressure, and will work with little or no further 
adjustment over a range of perhaps 100 pounds pressure. With 
feed water at about 100 deg. F, the same injector would start at 
30 or 35 Ib. steam pressure, and will work over a range of per- 
haps about 70 Ib. or up to about 100 Ib. 

In addition to the automatic injector there is another type 
having two sets of tubes, one for lifting and one for forcing. 
Such instruments are often termined inspirators to distinguish 
them from the ordinary automatic injector. When properly ad- 
justed the lifting set of tubes acts as a governor to the forcing 
set, supplying under a great range of steam pressure the proper 
amount of water to condense the steam in the final set of tubes. 
Such an injector handling cold water with a short lift, will work 
through a range of over 200 Ib., while with water as hot as 100 
deg. F and small lift it will work through a range of from 150 
to 200 Ib. The operation of each set of tubes is on the same 
general principles as above described for the automatic injector. 


The office of the feed heater is to raise the temperature of 
the feed-water from that of the hot-well as nearly to that within 
the boiler as may be practicable before the feed enters the boiler 
proper. Feed heaters are of two fundamentally different types, 
according as the heat used is drawn from waste furnace gases 
or from steam. If the former, as is very common with water- 
tube boilers, as described in Section 14, the feed heater is really 
a part of the boiler. The entire operation is thus performed in 
two stages, one in the heater, and one in the boiler proper. In 
the first stage it is sought to raise the water as nearly as possible 
to the boiling point by use of the furnace gases after they have 
passed the main steam generating tubes. In the second stage, 
carried out in the boiler proper, the previously heated water is 
transformed from liquid into vapor. 

The resulting economy comes from being thus able to re- 
duce the products of combustion to a temperature lower than 
they would otherwise have before finally getting rid of them. 
The addition of the heater will, of course, affect the draft, and 
the extent to which heating surface, either in the form of steam 


generating tubes or feed heating' tubes, can be added without 
seriously interfering with the draft is a point which must receive 
consideration. Morever, since the feed-heating surface might be 
put into additional main boiler tubes, thus giving the same total 
surface without a feed-heater, the question may naturally be 
asked whether in such case the results would be as good. In 
other words, is it better to put the total heating surface all into 
main boiler tube surface, or to divide it up and put a part 
(usually quite small) into a feed-heater located beyond the mail? 
part of the boiler? Experience seems to indicate the latter as 
the better design of the two, and the fundamental reason is that 
given above viz., that we are thus able to reduce the products 
of combustion to a lower final temperature than with main boiler 
tubes of the same aggregate surface. The reason for this is 
found in the fact that the temperature of the feed water as it en- 
ters the heater is much lower than that of the steam and water 
within the main tubes. Hence with such a heater the gases as 
they leave the boiler pass over relatively cool surfaces, and the 
flow of heat will be much more pronounced, and the gases will be 
more effectively cooled than by passing over an equal area of 
main boiler tube surface. 

Turning now to the other type of heaters we have an en- 
tirely different mode of operation. Here the heat given the 
feed-water comes from steam which is drawn either directly from 
the main or auxiliary steam pipe, or from the receivers, or from 
the exhaust of some of the auxiliaries on its way to the con- 
denser or to the escape pipe. There are two styles of heater 
working on this principle. In one the steam and feed-water are 
mixed together in the same chamber and the steam is condensed 
and thus joins the feed-water, raising its temperature as may be 
determined by the conditions of operation. In the other style 
the steam is on one side of a coil or nest of tubes and the feed- 
water on the other side, the heat passing through the metal of 
the tubes'from the former to the latter while the steam condensed 
in consequence of the loss of heat is drawn or trapped out as 
may be required. 

Where the steam and feed-water are mixed together the 
feed-heater consists essentially of a chamber or drum, as in Fig. 
191, provided with means for introducing the feed as a spray 
or series of cascades, while the steam is introduced in jets, and 
the two thus become intimately mingled. 


In the form here shown the feed enters through C and pass- 
ing out through a valve, D, falls as a cascade through the an- 
nular space between the pipe and the steam delivery drum which 
is pierced, as shown, with small holes. The steam enters 
through B, and passing through the holes in small jets becomes 
mingled with the water and thus imparts to it its heat. The t\Vo 
then fall to the bottom of the chamber from whence the feed 


Fig. 191. 


Feed Water Heater, Direct Contact. 

pump takes its supply, and by means of which the water is finally 
sent to the boiler. 

The advantages of such a form of heater are simplicity in the 
apparatus itself and a quicker action than with tubes, as in the 
other form. The chief disadvantage lies in the fact that an ad- 
ditional feed pump must be provided ; one for forcing the water 
from the hot-well to the feed-heater and the second for taking it 
from the heater and forcing it into the boiler. 



Turning to the other style of heater, as illustrated in Fig. 
1 92, we have a chamber or drum containing a nest of corrugated 
copper pipe. The feed-water passes on one side of the pipes 
and the steam on the other, as shown, the heat passing through 
from the one to the other as above described. Where such a 
heater is fitted to utilize the steam from an auxiliary exhaust, the 
heater forms simply a part of the exhaust passage and the steam 
passes through continuously, simply leaving a part of its heat 






Fig. 192. Feed Water Heater, Surface. 

In other forms of heater there is no continuous flow of steam 
through, but the steam is led to the steam side from the source 
selected and there gradually condensed by the loss of heat, while 
the water thus formed is drawn off as may be necessary. The 
heater is thus kept cleai for efficient operation, and the water is 
returned to the hot-well or otherwise into the feed system. 

In heaters of this type the flow of water is continuous from 
the feed-pump through the heater to the boiler, and no additional 
pump is required as with the other type referred to above. The 
difference is due to the fact that where the water and steam are 


separated, the water side of the heater can be operated under the 
full pressure in the feed-pipe and thus made part of the feed 
circuit. Where the water and steam are mixed the interior of 
the heater cannot be operated under any such pressure, and thus 
two separate pumps are required. 

In heaters using steam from the steam pipe or from the re- 
ceivers, it is clear that all such steam would ultimately go to the 
condenser and thence back to the boiler as feed-water, and hence 
that no heat is saved which would otherwise have been thrown 
away. In such cases it is not at first sight clear where the gain 
can come in, for the operation seems to be a simple shifting of 
heat from one part of the cycle to another without gain or loss in 
total amount. As a matter of fact the operation does consist of 
just such a shifting of heat, and here is where the gain in work 
comes in. This shifting of heat from one part of the cycle or 
routine of the steam to another introduces a change which brings 
the cycle a little nearer that for the highest efficiency as described 
in 59. While therefore there will be no saving of heat as such, 
the engine may be enabled to better use the heat which is pro- 
vided and thus to show a larger return in useful work. These 
points cannot be here discussed in detail, but it seems at least 
worth while to note in general terms the chief source of the 
economy experienced with such heaters. 

Especially will heaters of this type affect the routine of the 
steam favorably when they are arranged on the compound or 
step by step principle. In this arrangement the feed passes 
through a first chamber and receives heat from steam drawn 
from the low pressure receiver. It then passes on to a second 
chamber and there receives heat from steam drawn from 
the next higher receiver, and so on, in the last receiving 
a final addition of heat from steam of full boiler pressure. 
This type of heater, when of sufficient capacity to raise the tem- 
perature of the feed nearly up to that of the water in the boiler, 
will effect a marked economy in the engine, a saving presumably 
due to the change thus effected in the routine through which 
the steam is carried. 

In addition to the saving thus effected by feed heaters, many 
engineers believe that they are of use in reducing the strain and 
wear and tear on the boilers by furnishing a hot rather than a 
cold feed, and hence that they are of distinct advantage to the 
boiler and well as to the engine. 



Sec. 31. FH/TERS. 

Feed water filters are provided for removing oil from the 
feed water before it enters the boiler. See further on this point 
Section 41. Such filters are made in various forms, the chief 
features being the kind of filtering material employed, and the 
arrangement of the flow of water through it. Animal charcoal;, 
sand, gravel, broken pumice stone, etc., form one class of sub- 
stances, while fibrous materials such as sponges, bagging, towel- 
ing, etc., form another case. Of the first class, animal charcoal 
is the best, though somewhat expensive. It may, however, be 


Fig. 193. Feed Water Filter. 

removed from time to time, washed in lye water and replaced r 
and thus made to do duty for a long period of time. The various 
fibrous materials, such as sponges or bagging, soon become 
clogged also with oil and impurities, and require either replace- 
ment or washing and cleaning. After a few repetitions of 
cleaning in this manner, such material must be replaced with 

Filters also differ, according to whether the flow of water 
through the filtering material is forced by the pressure of the 
feed pump, or is due simply to the flow under the action of 
gravity. In gravity filters the flow may be up and down, two or 


three times through the bed of filtering material, the course be- 
ing determined by suitable partitions or passages within the filter 
box. When the action is under pressure the filter forms a part 
of the feed pipe circuit and the water enters and passes through, 
urged by the action of the feed pump, though at a much lower 
velocity than through the feed-pipe itself. In such case, if the 
filter should become choked, excessive pressure might be de- 
veloped between the feed-pump and filter with the possibility of a 
rupture of the latter. To avoid this danger a by-pass pipe and 
safety-valve may be arranged so that the valve will open under 
an excess of pressure and allow the water to flow around the 
filter to the continuation of the feed-pipe beyond. The safety- 
valve may also be maintained open by appropriate means and 
the filter shut off by stop-valves, thus sending the feed through 
the by-pass pipe and leaving the filter free for examination and 

In Fig. 193 a simple form of filter is shown with by-pass pipe 
and valves for controlling the flow of the feed. From the ex- 
planation above the operation of the filter will be readily under- 


The office of the evaporator is to supply fresh water to make 
up the loss in boiler feed. Of the steam which the boilers supply 
not all can find its way back through the feed. Small steam 
leaks may occur at the various joints and stuffing boxes, some of 
the auxiliaries may not send their steam to the condenser, the 
whistle may be used (as in foggy weather), and so in various 
ways losses of fresh water will occur. The proportion of such 
loss varies widely with the circumstances, but will often amount 
to 5 per cent, and more. In order to avoid making up this loss 
with salt water the evaporator is provided. 

A modern representative evaporator consists of a series of 
nests or coils of pipe contained within a chamber, as shown in 
Fig. 194. The chamber has a salt-water inlet, and steam from 
the boiler or from one of the receivers is passed inside the tubes. 
The heat in the steam passes through the tubes and forms steam 
or vapor of lower pressure on the salt-water side. The chamber 
is connected with the condenser or with the low pressure re- 
ceiver, and the steam formed in the evaporator is thus passed 
into the main circuit and serves to make up the loss as specified 
before. At the same time the water formed in the coils bv the 



loss of heat is drawn or trapped out as it accumulates, and is re- 
turned to the feed, so that all steam formed on the salt-water 
side is a net gain for the fresh-water account. The coils on the 
outer or salt-water side naturally become coated with scale, so 
that they must be cleaned from time to time. To this end they 


Fig. 194. Evaporater. 

are usually made removable or arranged so as to be readily 
accessible through manhole openings in the shell. It is, of 
course, essential that the tubes be kept clean for the most effi- 
cient working of the evaporator. 

In the operation of the evaporator the chief point requiring 
attention is the proper proportion between the amount and tern- 


peraturc of the inflowing steam and the pressure within the 
chamber on the salt-water side. If the pressure is low and steam 
is provided in excess, it may give rise to a violent ebullition or 
foaming, which will carry some of the salt water along with the 
vapor formed, and thus introduce salt into the circulating sys- 
tem. This condition must be guarded against by a proper con- 
trol of the amount of steam admitted. 

In the way of general maintenance, the tightness of the tube 
joints and the condition of the tubes as regards scale are the 
chief points requiring attention. 


In early marine practice the fly-wheel pump was a favorite 
type, and was used for all ordinary purposes where an independ- 
ent pump was required, as for boiler feed, fire purposes, or for 
general purposes on shipboard. This pump consisted essen- 
tially of horizontal steam and water cylinders with the piston and 
plunger on a common rod and moving together. Attached to 
the rod was a cross-head with connecting rod leading to a crank 
and shaft carrying a fly wheel. The fly wheel served to carry 
the pump past the dead points, and the shaft served to carry an 
excentric which actuated a simple slide valve on the steam 

This type of pump, however, has almost entirely disappeared 
from modern practice, its place being taken by the direct acting 
pump with its greater compactness of form and better adapta- 
tion to the conditions of service. 

We will now consider briefly the essential features of this 
type of pump, with a few examples drawn from modern practice. 

As illustrated in Fig. 195, the pump is horizontal and con- 
sists of two cylinders, one for steam and one for water, carried 
on a common piston rod. The steam end is operated by means 
of a suitable valve gear as a simple reciprocating engine, and 
thus communicates the same movement to the pump plunger or 
piston. Each end of the water cylinder is provided with both 
inflow and outflow valves, as shown, and thus the pump becomes 
double acting, that is delivering on each stroke alternately from 
one end and the other. 

For operating the steam ends of pumps of this type, a great 
variety of ingenious valve gears have been devised. 

The need for special device arises from the fact that there is 


no rotating part and no chance to use an excentric, and that the 
valve cannot be operated directly from the main piston rod. 
Where it is thus required that a single set of principal moving 
parts be self operating, the valve gear usually consists of the 
following chief features : 

(i.) The main steam valve, often of special form, but usu- 
ally operating as a simple slide valve. 

(2.) An auxiliary plunger or piston moving in a cylinder 
formed in the valve chest, and coupled or connected to the main 

(3.) An auxiliary valve controlling steam and exhaust to 
and from the two ends of the auxiliary plunger cylinder. 

(4.) Means for operating the auxiliary valve from the main 
piston-rod. Such means may consist of levers, links, rods, cams, 
etc., operated by tappets on the main rod, whose location or 
point of operation may be adjusted according to the length of 
the stroke desired. 

The chain of operation is then in general as follows : 

Just before the end of the stroke the tappet or other piece 
moved by the main piston rod gives motion to the auxiliary 
valve. This produces an adjustment of steam and exhaust for 
the auxiliary cylinder which results in a movement of the auxil- 
iary piston and hence the movement of the main valve as de- 
sired. The motion of the main piston is thus reversed and the 
stroke takes place in fhe opposite direction, and so on continu- 

In Fig. 195 the lower section is horizontal and taken through 
the auxiliary piston and auxiliary slide valve operated by the 
levers and links as shown. The upper view shows in vertical 
section the auxiliary piston and main steam valve. 

If two such pumps are placed side by side, it is found that 
the valve of each pump may be operated from the piston rod of 
the other. Hence by appropriate connections a pair of such 
pumps may be made self operative, the strokes being made alter- 
nately, and each piston rod running the valve gear of the other 
pump. Such an arrangement constitutes a dnplc.v pump, a form 
which has enjoyed w^ide and continued favor among marine en- 
gineers for feed-pumps and for other purposes with generally 
similar conditions. 

In the so-called "Admiralty" style of pump the motion of 
these parts is vertical, and the water valves are specially arranged 



with a view to ready examination and overhauling. As noted in 
Sec. 29, this general style is quite commonly used for feed-pump 
purposes. Such pumps may be either simple or duplex, but the 
duplex type is more commonly met with in this form. In Fig. 
196 is shown one member of a duplex admiralty pump, the ar- 
rangement of the parts and operation of which will be apparent 
without further explanation. In Fig. 197 is shown similarly a 
single vertical type of feed-pump with independent valve gear, 

Marint E.iytnterfng 

Fig 19"). Direct Acting Independent Feed Pump. 

the auxiliary piston being operated by the opening and closing 
of ports due to a rocking motion which is communicated to it by 
the levers and link work as shown. 

Bilge pumps when independent, and all general service 
pumps, are usually of the direct acting form as above illustrated. 
The chief item of difference is found in the ratio of the areas of 
steam piston and water plunger. Where the water is to be de- 



livered under considerable pressure, as for feed-pumps or for fire 
purposes, the area of steam piston will be from two to three 
times that of water plunger. Where the resistance to be over- 
come is less, as in a pump for freeing the bilge or for circulating 

Fig. 196. Vertical Duplex Feed Pump. 
Admiralty Type. 

Marine Engineering 

Fig. 197. Vertical Single Feed 

water through distillers, evaporators, water-closets, etc., the 
water plunger may be relatively larger and we shall find such 
pumps with a water end only slightly smaller or equal in size or 
even larger than the steam end. 


Such pumps are always so connected up, of course, as to 
enable them to be run from the auxiliary boiler. 


The centrifugal blower is the type universally used on ship- 
board for all purposes requiring the handling of large quantities 
of air under light pressure, as for ventilation and forced draft. 

As indicated in Figs. 95, 96, such a blower consists of a 
series of flat or nearly flat steel vanes carried on a shaft and sur- 
rounded by a casing. The principle of operation is the same as 
with the centrifugal pump, as described in Sec. 26. The rotation 
of the vanes sets up first a circular current or rotation of the air, 
and as a result of this motion, centrifugal force is developed 
which carries the air out toward the tips of the blades and de- 
velops an increase of pressure from the hub outward. If an out- 
let is then provided in the outer shell of the casing the air will be 
delivered at this point and the surrounding air will flow in to 
take its place at the intake about the hub. So long as the rota- 
tion is kept up these conditions will continue, and there will be a 
continuous flow of air in at the hub and out through the delivery 
passage under a pressure depending on the speed of rotation and 
other circumstances. 

Blowers are driven by either steam engine or electric motor 
direct connected to the shaft, and are made of various forms so 
as to readily find a place in almost any position desired, thus re- 
quiring the smallest possible amount of otherwise valuable space. 

In the operation of blowers the points of chief importance 
relate to the general care which must be given to the operating 
motor, whether electric or steam, and to the proper lubrication 
and care of the fan-shaft bearings. 


In many types of water-tube boilers, special arrangements 
are provided for separating the steam from the water. These 
are usually located in the upper drum or chamber and consist 
commonly of one or more metal plates pierced with holes 
through which the steam passes to the stop-valve and steam 
pipe, and which exercise more or less of a straining or separat- 
ing action on the water and steam. Reference has been 
made in Sec. 16 [4] to arrangements of this character. 

In addition to such arrangements located in the upper drum 
of water-tube boilers, special devices known as separators are 



used wherever the steam is likely to have any considerable pro- 
portion of water. Such devices are found in great variety of 
form, and utilize various principles in their operation. The most 
successful are those which employ for separating the water 
from the steam the centrifugal force developed by a rotation 
of the steam as it enters or passes through the separating 

The .following description will serve to illustrate the oper- 
ation of a typical separator of this character : 

The separator, as shown in Fig. 198, consists of a vertical 
cylinder with an internal central pipe extending from the top 
downward, for about half the height of the apparatus, leaving an 
annular space between the two. 

e Engineering 

Fig. 198. Separator. 

A nozzle for the admission of the steam is on one side, the 
outlet being on the opposite side or on top as may be most con- 
venient in making the connections. 

The lower part of the apparatus is enlarged to form a re- 
ceiver of some considerable capacity, thus providing for a sudden 
influx of water from the boiler. 

A suitable opening is tapped at the bottom of the apparatus 
for a drip connection, and a glass water gauge shows the level of 
the water in the separator at all times. 

The current of steam on entering is deflected by a curved 
partition and thrown tangentially to the annular space at the side 



near the top of the apparatus. It is thus whirled around with the 
velocity of influx, and a centrifugal force is developed, which 
throws the particles of water against the outer cylinder. These 
adhere to the surface, so that the water runs down continuously 
in a thin sheet around the outer shell into the receptacle below, 
while the steam, following a spiral course to the bottom of the 
internal pipe, enters it abruptly, and in a dry condition passes 
upward and out of the separator, without having once crossed 
the stream of separated water, all danger of the steam taking up 
the water again after separation being thus avoided. 

The water thus separated from the steam collects in the 
lower part of the chamber and may be drawn out from time to 
time or it may be led to a steam trap of approved form and 

Fig. 199. Ash Ejector. 

trapped out, thus making the operation entirely automatic. The 
water thus obtained will, of course, be of high temperature and 
should be led directly to the hot-well where it will aid in raising 
the temperature of the feed-water. The heat which it contains 
will thus be returned to the boiler, and saved, and all heat loss 
in connection with the operation will be avoided. 

Sec. 36. ASH EJECTOR. 

Ashes are either hoisted in a bucket by a special hoist to a 
point on the main deck level and there dumped into an ash- 
chute leading to the side of the ship and down into the water, or 
else disposed of directly* from the fire-room by means of an ash 
ejector. Such a device is illustrated in Fig. 199. A represents 
a cast metal chute or pipe leading from the fire-room up and out 



through the side of the ship near or slightly above the water line. 
At the lower end this chute connects with a hopper, B, into 
which is led a pipe from the discharge of a pump. This pipe 
enters to a point near the lower end of the chute, into which its 
discharge is directed, and is contracted to a nozzle so that the 
water issues with a high velocity. The hopper may be closed by 
a cover, and if in this condition the discharge valve is opened and 
the pump started, a stream issues with high velocity from the 
upper and open end of the chute. If then the cover is removed 

Fig. 200. Ash Gun. 

and ashes shoveled into the hopper, they are caught by the 
stream, carried rapidly up, and ejected free of the ship's side in a 
mingled jet of ashes and water. 

It is found by experience that at the upper bend the wear 
on the metal of the pipe, due to the scouring action of the ashes, 
is very rapid, and it is usually found necessary to make this 
bend in a separate piece of extra thick metal, and to provide 
by means of a proper arrangement of joints for its replacement 
as occasion may require. 



A similar device known as the ash gun is shown in Fig. 200. 
Where possible the lead of pipe from the hopper to the ship's 
side is made straight .so as to avoid all bends and elbows. The 
principles of operation are the same as above explained. 

012 8 4 S 67 8 10 

48 Murin* Engineering 

Figs. 201, 202. General Arrangement Plans. 


We shall not here discuss in detail the various questions 
which may arise in connection with the problem of the general 
arrangement of marine machinery. It will be sufficient for our 


present purpose to note the fundamental principles which must 
be held in view : 

(i.) Each piece must be located so as to favor, as far as 
possible, examination and repair. 

(2.) Each piece should be located with reference to handi- 
ness of care and control in routine operation, and in such way 
as to interfere to the smallest practicable extent with the routine 
care, examination and repair of other pieces. 

(3.) Due regard must be had to economy of space and such 
combinations of the various pieces must be sought as will re- 
quire the minimum total space, while giving the necessary free- 
dom in accordance with the principles noted above. 

(4.) The influence of the location of the various pieces on 
the arrangement of the piping must be carefully noted and due 
weight must be given to simplicity, shortness and directness of 
the various lines of piping. 

In Figs. 201, 202 are shown illustrations of a general ar- 
rangement plan in which the condenser is located in the engine 
framing. The remaining features are independent, and include 
those most commonly met with in the auxiliary equipment of the 
engine room. 





In the present section it is the purpose to give brief hints 
and suggestions regarding the routine of operation and manage- 
ment in the fire-room in getting under way and on the voyage, 
first supposing that everything is working smoothly and with- 
out trouble, and then to notice the chief emergencies which may 
arise. We shall first suppose that fire-tube boilers are in use, 
and later give such supplementary suggestions as may be suit- 
able for water-tube boilers. 

[i] Starting Fires and Getting Under Way. 

A general examination must first be made of the boiler and 
fire-room equipment in order to make sure that everything is in 
readiness for getting up steam. Among the more important 
points to be attended to the following may be mentioned : 

See that the coal bunker doors are in proper working order 
and if the bunker is partly empty it may be well to air it by open- 
ing the door and taking off the deck plates. 

See that the coal handling gear is on hand and in proper 

See that the necessary fire tools are provided, and dis- 
tributed as needed. 

Examine the grate-bars, bridge-walls and back-connections, 
and note whether the area of passage above the bridge-walls is 
properly proportioned. For usual conditions it should be from 
1-5 to 1-7 the grate area. 

Note the condition of the tubes both from the front and 
back connections. 


Examine the dampers in uptakes and funnel to see if they 
are in working order, and open them preparatory to lighting 
the fires. 

Examine carefully all valves, cocks, piping and connections 
and make sure that everything is connected up as it should be, 
and that no valves are open which should be closed nor closed 
which should be open. 

See that no waste or other inflammable substances have 
been left about by workmen on the tops of the boilers. 

If the water has not been previously run up in the boilers, 
this may be under way in the meantime. In modern practice 
the boilers are always filled with fresh water where possible, ob- 
tained from a hydrant on the dock or water-boat alongside, and 
put in usually by a hose through an upper manhole. If, how- 
ever, the boat is lying in fresh water, or if by necessity the water 
is to be taken from overboard, it is then run in through the bot- 
tom blow and Kingston valve. In the meantime examine the 
connections leading to the water-gauge and cocks. Clean the 
glass if necessary, and make sure by means of a wire that the 
opening through the cocks is clear. The packing of the gauge 
glass should also be examined and renewed if necessary. When 
the water appears in the gauge glass and shows from one-half to 
two-thirds full in each boiler, it may be shut off. 

All open manholes may then be closed, and the boilers are 
ready for the fires. 

Notice of the time when steam is required should have been 
given not less than from six to eight hours in advance, and 
many engineers prefer a still longer time in which to bring along 
everything into working condition. With hard coal a certain 
amount of wood is necessary in starting the fires. With soft 
coal less wood is required, and if necessary oily waste may be 
made to answer the purpose. If fires are up in the donkey 
boiler a little live coal may be taken from them to assist in start- 
ing. As soon as the fires are going the hydrokineters are put on 
if such appliances are fitted. In some cases arrangements are 
made for drawing the water by the donkey or auxiliary feed- 
pump from the bottom of the boiler by the bottom blow and re- 
turning it through the feed-pipe, thus producing a forced or as- 
sisted circulation. Where there are no appliances for forcing the 
circulation during this period, it is considered good practice to 
light first the fires in the center furnaces, and later, by one or 


two hours, those in the wing furnaces. The natural circulation 
thus produced will more nearly even up the temperature within 
the boiler than if all fires are lighted at the same time. After 
the fires are fairly going the funnel or uptake dampers may be 
partly closed so as to hold the fires back, and bring them along 
at a moderate gait as desired. 

While the boilers are thus warming up and before steam has 
formed, a last look may be given to the boiler mountings and 
their connections. The various cocks and valves should be 
worked, and especially the stop and safety valves, in order to 
make sure that none are jammed or in any way out of order. 
The oil lamps for the steam and water gauges may also be 
trimmed and lighted, or the electric bulbs cleaned, if such are 

During this period the steam-pipe drains and safety valves 
are usually open to allow of the escape of the air and of the con- 
densed vapor as formed. In some cases, however, the safety 
valves are closed, and the stop valves being open, the air and 
vapor are expelled along the steam-pipe and through the engine, 
thus beginning the process of warming up. Many engineers, 
however, prefer to keep the boiler stop valves closed until steam 
is formed, and to discharge the air through the safety valve, or 
in some cases through a specially fitted air-cock. If steam is 
already up on some of the boilers or if there is no auxiliary 
steam-pipe and the pressure from the donkey boiler is on the 
main steam-pipe, then of course the stop valves must be closed 
on the boilers in which steam is being raised, and they must re- 
main closed until the pressure on the boiler is equal to that in the 
steam-pipe. In opening a boiler stop valve connecting with a 
pipe in which there is no pressure the following precautions 
should be taken : 

(1) The pipe should be thoroughly drained and especial 
care should be taken that there are no sags, bends or U's un- 
provided with proper drains, and in which a pocket of water may 
have collected. 

(2) The valve should be very carefully eased from its seat 
and opened only from a quarter to a half turn until the pipe is 
under full boiler pressure and has taken the temperature of the 
steam, and the drains are discharging steam instead of water. 
In opening a boiler stop valve connecting with a pipe in which 
there is approximately the same pressure as in the boiler, it is 


simply necessary to ease the valve from the seat and note by the 
sound whether there is a sufficient difference of pressure to cause 
any violent flow in one direction or the other. As soon as the 
absence of such evidence indicates an equality of pressure on 
both sides of the valve, it may be opened out as desired. 

The two fundamental principles underlying much of this 
routine and detail are simply as follows : 

(1) To prevent as far as possible sudden changes in the 
temperature condition of the boilers, piping and machinery, and 

(2) To prevent throughout the steam-pipe system the accu- 
mulation of water at any point. 

If these two points are kept clearly in view and good en- 
gineering judgment used in carrying them out, the life of the 
boilers and machinery will be prolonged, and danger of ruptured 
pipes through the effects of water hammer will be avoided. 

After steam is formed and the pressure has risen to some 40 
or 50 pounds the hydrokineters may be shut off, especially if the 
ship is to get under way as soon as ready. If, however, the boil- 
ers are to stand some time with steam up, it may be advisable to 
turn on the hydrokineters from time to time, at least as long as 
the pressure in the donkey boiler is sufficient for the purpose. 

The fires in the meantime have been kept simply in good 
condition without forcing, and even if they work under a forced 
draft system, only enough air should be provided during this 
stage to bring them along at the gradual pace which will allow 
the boiler to properly accommodate itself to the change in tem- 
perature and other conditions. 

The fire-room auxiliary machinery should also be examined 
during this period, and tested under steam from the donkey 
boiler if possible. The feed pumps should first receive attention, 
in order that there mav be no question as to the proper supply of 
feed-water when required. 

The fan engines should be examined, oiled and turned over 
under steam. 

The ash-hoist gear and engine, or ash ejector and pump, 
should be examined and put in working order. 

If steam for these purposes is not to be had from the donkey 
boiler, then as soon as a sufficient head is formed on the main 
boilers these auxiliaries must be examined, taking in all cases the 
feed pump first. 

Soon after lighting fires it may be desirable to slacken up 


somewhat on the funnel guys on deck, in order that the expan- 
sion of the funnel may not bring an undue stress upon the guys 
and their connections, or upon the funnel and its supports. 
After the ship is away and the funnel has taken its temperature 
for running conditions, the guys may be tightened up so as to 
properly support the funnel in a sea way. 

[2] Fire Room Routine. 

At length we may suppose that full head of steam has been 
formed on the boilers, that the fires have been brought up to 
proper condition, and that the ship has gotten under way for the 
voyage. As soon as possible the operations in the fire-room 
should be brought to a regular routine. This will involve the 
following chief features, which we shall consider separately: 
(i) Firing. (2) Water tending. (3) Disposal of ashes. 
(4) Cleaning fires. (5) Sweeping tubes. 

Firing. The routine of firing should be so arranged that no 
two furnaces in boilers connected to the same stack shall be open 
at the same time. If this is not practicable, then care must be 
taken to avoid at least firing at the same time furnaces in oppos- 
ite ends of double-end boilers, especially if there is a common 
combustion chamber. Two furnaces in a single-end boiler, or in 
the same end of a double-end boiler will, of course, never be 
fired at the same time. It is now well understood that firing 
light and often is better than heavy and at great intervals. 
There is, however, a limit to this, for the oftener the firing the 
more are the furnace doors open and the more is the draft subject 
to disturbance, while the arrangement of a suitable routine be- 
comes more and more difficult. 

Light and frequent firing, especially with water-tube boil- 
ers, is now, however, the rule where the best results are to be 
obtained. The furnace door should be opened smartly and kept 
open only the minimum time needed to get the coal on. Hard 
coal is spread in as even a layer as possible over the grate. For 
firing soft coal two methods are available. When firing for coal 
efficiency, that is to get the most heat out of a pound of fuel, 
the coal should be first charged in front and coked, and then 
should be pushed back and burned. When firing for weight 
efficiency, that is to get the most power out of the boiler, the 
former method would be too slow and the coal must be spread 
over the fire and burned without waiting for the separate dis- 


tillation and combustion of its gases. Where the coal runs ir- 
regular in size the large lumps should be broken into pieces not 
larger than the fist. The thickness of the fires varies with the 
conditions, from six to ten or twelve inches, or even thicker un- 
der a heavy forced draft. With a given speed of fan the air 
pressure in the ash-pits will vary widely with the thickness of 
the fire, rising as it is thicker, and falling as it is thinner and the 
air finds more ready passage through. With a thick fire it will 
therefore be easy to get a strong draft pressure in the ash-pits, 
while with a thin fire it will be perhaps impossible, even with a 
much higher speed of fan. A strong draft pressure will not, 
however, produce the corresponding rate of combustion unless 
the thickness and condition of the fire are such that the pressure 
is able to drive through it the necessary amount of air. For the 
best combustion the thickness of the fire should be so adjusted to 
the draft pressure that the latter is able to drive the necessary 
air through, and keep it in a state of active combustion through- 
out from fire grate to upper surface. Care must be taken to 
keep the grates evenly covered, especially at the back, and to 
prevent the formation of relatively thin or bare spots. A spot 
which is relatively thin allows of the passage of relatively more 
air. This further increases the combustion at that point and 
the spot becomes still thinner, thus allowing more and more air 
to escape freely instead of passing through the remainder of the 
fire as it should. 

In the intervals of firing the pricker and slice bars may be 
used to clear away the ashes and clinker, if such is forming. 
Care should be taken to prevent the formation of dull or dead 
spots due to "the accumulation of ashes or clinker, especially at 
the corners of the grate. Among old firemen a familiar saying 
relating to this point is : "Take care of the corners and sides 
of the fire and the middle will take care of itself." The ash-pits 
should also be kept clear of ashes, for if allowed to accumulate 
they will prevent the passage of air to the grate, especially at the 
back. If the passage of air is thus interfered with to any con- 
siderable extent there will be also danger of overheating the 
grates and of bringing them down into the ash-pits. 

In connection with the use of the slice bar, it must not be 
forgotten that every opening of the furnace door means an in- 
rush of cold air into the furnace, a checking of the draft, a distur- 
bance of the combustion, and often severe strains on the struc- 


ture of the boiler, due to the sudden chilling and contraction 
which the heating surfaces undergo. If shaking grates are fitted 
much of this cleaning may be done without opening the door, 
though no form of grate is quite able to deal satisfactorily with 
coal showing a decided tendency to form clinker. 

In thus working the fires a certain amount of fine unburned 
or partly burned coal will be shaken down into the ash-pit. In 
some cases this forms so large a part of what comes through the 
grate, that it may be immediately thrown onto the fire and 
burned over again. In most cases a sifting or washing of the 
ashes and separation of the combustible from the non-combus- 
tible would show a surprisingly large per cent, to be available as 
fuel, and some saving could usually be effected in this way. It 
is rare, however, that anything of the kind is attempted, as with 
present prices of coal it. may be doubted whether the additional 
appliances and labor would be paid for by the saving effected. 

Water-tending. The care of the water is the most important 
and responsible of the duties in the fire-room. The ideal is to 
keep the water regularly flowing inward through the check- 
valves at about the same rate as it is flowing outward as steam 
through the steam pipe. This requires constant watch and ad- 
justment of the valves, closing down where it is entering too rap- 
idly and opening up where it is entering too slowly. Instead of 
this method it is often the custom to put the feed on strong first 
to one boiler and then another, in order, according to the firing, 
feeding the boiler up when the fire is at its best, and shutting 
down when it is freshly coaled. The steady and uniform feed is, 
however, better because it approaches more nearly to a uniform 
condition of the boiler, especially on the water side. 

The position of the water is determined, of course, from the 
water gauge and cocks. It is necessary, of course, that the 
gauge and its connections be clear of any obstruction in order 
that the height of the water may be properly indicated. To make 
sure that everything is clear the gauge glass and connections 
are blown through by the "double shut off" method as follows : 
In Fig. 203, G represents the glass, A the drip cock, B and C the 
cocks connecting to the stand, and D and E those connecting 
the stand to the boiler. First, B and E are closed, and A is 
opened. If steam blows through it shows that A G C D are 
clear. Second, C and D are closed and A is opened. If water 
blows through it shows that ABE are clear. The action of the 



water in the glass will usually show to an experienced eye 
whether or not the connections are clear. If the water is lively 
and follows the rolling of the ship it is a good indication that the 
passages are clear. Otherwise it indicates that an obstruction 
exists which must be sought out without delay. In the mean- 
time the water cocks are relied upon, and in fact many ex- 
perienced water tenders prefer the indications of the cocks to 
those of the glass, while they should in all cases be freely used 
as a check on the glass. To those without experience, however, 
the glass is less apt to be misleading. The indications of the 

Fig. 203. 

Marine Ziiyizeeriny 

Test for Water Gauge and Glass. 

water cocks are sometimes difficult to interpret, because fre- 
quently it is not easy to tell whether water or steam is blowing 
off. With high pressure steam especially, a jet of water issuing 
at a temperature of 350 degrees to 400 degrees is instantly sur- 
rounded with a shell of vapor formed by the vaporization of part 
of the jet. Furthermore, if the water in the boiler is in active ebul- 
lition near to the surface so that the jet would be drawn from a 
mixture of steam and water, then on issuing it becomes practical- 
ly a jet of moist steam. On the other hand if the water is well 
below the cock so that the jet would be drawn from steam alone 


or from moist steam, then on issuing it will become dry and 
usually super-heated. It is also a fact, especially with water-tube 
boilers, that due to a kind of lifting action, a cock will often dis- 
charge moist steam or a mixture of water and steam, even if the 
water level is somewhat below the mouth of the cock. It is 
hence readily seen that the indications from the cocks must be 
interpreted with judgment, and that some experience is neces- 
sary in order to always draw correct conclusions from them. 
It is often difficult to distinguish between an empty glass and 
one entirely full. In order to make sure close the cock B, Fig. 203, 
and slightly open A. If the gauge is full of water the surface 
will gradually descend, first coming into view in the top of the 
glass and then passing out of view at the bottom. If then A is 
closed and B is slightly opened, the water will rise again in the 
glass and pass out of view at the top. 

Blowing off. Blowing off boilers to reduce the concentra- 
tion or density of the water is now rare in good practice. In- 
stead of reducing density by introducing sea-water for feed make 
up, evaporators are installed for providing fresh water make up, 
or for short runs additional fresh water is often carried in double 
bottoms or spare tanks provided for the purpose. 

In modern practice the purpose of blowing off is (i) to get 
rid of mud or slush in the bottom of the boiler or in the special 
mud drums of a water-tube boiler, and (2) to get rid of oil and 
scum at and near the surface of the water. For the former a 
bottom blow or special mud cock is required, while for the latter 
the surface blow is used. In ordinary experience on deep sea 
voyages where evaporators or other fresh water make up is pro- 
vided, the use of the surface blow is all that is needed to keep 
the water in good working condition. It must not be forgotten 
that the use of the blow-off means a direct loss of heat, and hence 
it should be used with discretion, and no more frequently than is 
needed for the purpose in view. An idea of the condition of the 
water in the boiler near the surface may be obtained by drawing 
off a little water from a cock fitted into the surface blow pipe, or 
from a gauge cock fitted directly to the boiler. The water being 
allowed to cool and settle, is then poured into a glass jar, when 
its condition is readily noted, and the need of using the blow 

It may be well to speak at this point of the proper method of 
testing, from the outside, the correct position of a plug cock han- 


die for "closed" and for "open." Instances have been known 
where there was no mark on the head of the plug, and the han- 
dle becoming bent or being wrongly placed, the cock was left 
shut when it was supposed to be open, or open when supposed 
to be shut, with the possibility of most serious consequences, 
especially in the latter case. 

A careful examination of the cock, aided if need be by 
placing the ear to the chamber, will suffice to tell whether or not 
the cock is open and water or steam passing through. Then the 
cock being open let it be turned in one direction until it is just 
closed, and then back in the other direction until it is closed 
again. Half way between these two positions it will be wide 
open, and at right angles to the latter position it will be full shut. 

Taking the Saturation or Testing the Density of the Water. 
The density of the water is determined by the use of a hydro- 
meter or salinometer as it is often termed. Under modern con- 
ditions where evaporators provide fresh water make up, the 
density rises but slowly, and it is usually only necessary to ob- 
serve its value once or twice a day. It is usually not allowed to 
rise above two or two and one-half. See Sec. 17 [10] . 

Disposal of Ashes. For the disposal of ashes two chief 
means are in use. According to the older method they are sent 
up and disposed of through an ash chute leading overboard and 
down the side of the ship to the water, and this, method is still 
extensively used in large and deep ships. In the more modern 
system they are disposed of from the fire-room direct by means 
of an ash ejector. In either system it is usually sufficient to 
dispose of the ashes once in a watch, and they are collected, wet 
down and either hoisted in buckets or shoveled direct into the 
ash hopper usually between 6 and 7 bells. 

Cleaning of Fires. The routine working of fires spoken of 
above will suffice to keep them in fairly good condition for sev- 
eral hours, provided the coal is of fairly good quality. It usually 
becomes necessary, however, to give to each fire from time to 
time a more thorough cleaning than is possible in the manner 
previously referred to. To this end the fire should be taken 
when partly burned down, but not too far lest there be nothing 
left after cleaning on which to build up again. One side may 
be cleaned first, working the good coal over to the other side, 
separating out the clinker and ashes, and hauling out the latter. 
Then similarly with the other side, working the good coal over to 


the side first cleaned and pulling out the clinker and ash. The live 
coal is then spread over the grates, fired lightly, and so brought 
up again into regular conditions. In some cases there is so little 
left after a thorough cleaning that live coal must be borrowed 
from another furnace to save the fire. Only judgment and ex- 
perience can determine the best point at which to clean a fire so 
as to insure the minimum loss of heat, and at the same time 
have enough coal left to nicely build on again. Some engineers 
prefer to burn the fire almost completely down to the ashes and 
clinker, and then pull the entire contents of the grate out and 
start afresh. This method, however, chills the furnace and more 
seriously interferes with the operation of the boiler, and is not to 
be recommended. It must of course not be forgotten that heat 
is lost with the clinker and ashes withdrawn, and the general 
operation should be so conducted as to keep this loss down to 
the minimum possible. 

Under usual conditions the fires will need cleaning in this 
way at intervals of from 12 to 16 hours, or at least once a day. 

Sweeping Tubes. In addition to the cleaning of fires the 
tubes will require cleaning from time to time, dependent on the 
character of the coal and other circumstances. With soft coal 
and moderate draft they will soon become partly filled with 
soot and ashes, thus choking the draft still further, and prevent- 
ing the transfer of heat to the water through the metal of the 

To prepare for sweeping tubes the draft is checked, ash-pit 
doors put up, furnace doors opened and front connection doors 
raised. Care should be taken to wait until the fire is burned 
partly down before doing this, so that the circulation of air 
through the grates may not be shut off while the fires are too 
heavy, thus endangering the grate bars. For cleaning the tubes 
the ordinary wire tube brush may be used. This consists of a 
mounting carrying wire bristles and fitted usually with a jointed 
handle by means of which it is pushed and pulled through the 
tubes, thus cleaning out the soot and ashes collected there. 
A more modern method consists in blowing through tfre tubes 
with a steam jet. The mounting of this appliance consists of a 
flange or conical ring fitting closely to the end of the tube and 
provided with a steam nozzle directed along the center of the 
tube. A handle is provided for holding and guiding the ap- 
paratus, and steam is led to it by means of a flexible hose. By 


this means the ashes and soot are driven out of the tubes into 
the combustion chamber. By still another form of apparatus the 
jet is not directed into the tube but across the front end pro- 
ducing a suction, and thus drawing the ashes and soot to the 
front connection and discharging them up the funnel. 

The operation of sweeping tubes is one that is necessary to 
maintain the continued efficient operation of the boiler, but it 
must not be forgotten that it involves a serious disturbance to 
the draft of the whole battery, that the chilling of the heating sur- 
faces and interruption to the regular routine are hard on the 
boiler itself, and that hence, it should only be done when neces- 
sary and then as quickly as possible. 

Stopping Suddenly. With everything going along its regu- 
lar schedule, suppose that the engine is suddenly stopped. The 
dispositions to be taken w r ill depend on whether the stoppage is 
momentary or whether it is expected to last for some time. Here" 
again the caution regarding a sudden change in the conditions 
must be kept in mind. If the stop is but momentary it wifl 
probably be sufficient to shut off the draft, close the dampers and 
put on the feed strong, standing by to ease open the safety 
valves in case the pressure rises too near the point of blow- 
ing off. If the stop is to be longer it may be necessary to still 
further check the fires by putting up the ash pit doors and open- 
ing the furnace doors. Caution must be exercised in thus check- 
ing the flow of air through the grates lest there be danger of 
overheating the bars, or even of bringing them down into the abh 
pit. Of these various steps for checking the fires the opening 
of the furnace doors' and the sudden chilling of the heating sur- 
faces is the most objectionable and should not be resorted to 
unless necessary. As an additional means the fires may be 
freshly coaled, especially with dampened coal. This will check 
the formation of steam and provide fuel for bringing them into 
good condition for the next start. A period of stoppage like, 
this may also be taken advantage of to clean such fires as may 
be in need of it. 

In addition to checking the formation of steam, that which 
is formed may often be used in a variety of ways. If evaporators 
are provided it may be turned on to them and thus go toward 
increasing the store of fresh feed water. The bilge pump may 
also be put on strong, and if its exhaust is saved there will be 
no loss of fresh water. In some cases with independent air and 


circulating pumps a bleeder is provided for taking the steam 
direct from the main steam pipe to the condenser. Here it is 
condensed and then sent by the feed pumps back to the boilers, 
thus avoiding blowing off at the safety valves and the loss of 
fresh water, and allowing the fires to be gradually reduced to the 
condition desired for the period of stoppage. 

Here again in all of these operations general principles aro 
often worth more than a multitude of minor directions. These 
principles are (i) Sudden chilling of the boiler heating surfaces 
must be avoided as far as possible. (2) Fresh water in the form 
of steam should not be wasted, and (3) Care must be taken not 
to allow the grate bars to melt down. 

So far as relates to the general securing of the machinery 
and gear in the fire-room, the hints given in connection with 
getting under way will be a sufficient guide in reversing the 

Supplementary Hints Relating to Water. Tube Boilers. In 
water-tube boilers the circulation is usually more nearly natural 
than in fire-tube boilers, and circulating devices are not, there- 
fore, required. Steam may be raised in such boilers in from 
twenty minutes to one hour, depending on the type, character 
of the draft, etc. With this type of boiler it is especially neces- 
sary that for the best results the firing be light, often and regu- 
lar, and that the fires be kept as nearly as possible in a uniform 
condition. It is also necessary that the feed be regular, and 
the water must be carefully watched, since from the small amount 
contained, any lack of feed in a given boiler will be followed by 
rapid lowering of the level, and by a rapidly increasing likeli- 
hood of danger to the tubes. In water-tube boilers it is es- 
pecially necessary that nothing but fresh water be used as feed, 
and great care must be taken to keep the condenser tight and 
the fresh water make up ample in quantity. 

The tubes of water-tube boilers become coated with soot and 
ashes on the outer or fire sjfie, and it is usually a very difficult 
matter to satisfactorily clean them without the use of a steam 
jet. In continued steaming for long periods, it will usually be 
found necessary from time to time to let the fires die down some- 
what and to use what methods are available for blowing off and 
dislodging the soot from the tubes. 

When stopping or standing by, the same general means may 
be adopted as in fire-tube boilers. As regards injury through 


sudden change of temperature, the water-tube boiler is some- 
what less liable than the fire-tube. This is due to the nature of 
the construction which, especially with curved or built up ele- 
ments is much more elastic than in the fire-tube boilers. It is 
always better, however, to avoid sudden temperature changes 
where possible, and the same principles may be properly applied 
here as previously discussed in reference to the other type of 

Coming Into Port. When coming into port notice will 
usually be given some hours in advance, so that the fires may 
be worked into a condition in accordance with their expected 
disposition after arrival. If they are to be drawn and the boilers 
opened up for examination and repairs, they will be allowed to 
burn down as low as possible so as to use no more fuel than 
necessary, and to leave as small an amount as possible to finally 
haul, while at the same time sufficient steam must be maintained 
to bring the ship safely to her anchorage or dock. If, on the 
other hand, the fires are to be banked, they will not be allowed 
to burn so low. It may be recommended to bank fires on the 
front of the grates, as in such case the air is heated as soon as 
it enters the furnace and the boiler is kept at a more nearly even 
temperature than if they are banked at the back of the grate. 
As the fires are banked they should be cleaned and enough 
fresh coal put on to hold them in the condition desired. If the 
fires are properly managed there will be little extra steam after 
the engines are stopped, and this may be readily disposed of by 
means of the evaporator, bilge pump, bleeder, safety valve, etc. 
Loss of fresh water at this time is of course less objectionable 
than when on the voyage, and if desired the steam may all be 
blown off through the safety valves. Many engineers, however, 
object to using the safety valves and escape pipe for this purpose 
except as a last resort, and prefer other means as mentioned. In 
passenger vessels the noise occasioned is usually considered ob- 
jectionable, though to obviate this a muffler is frequently fitted 
in the escape pipe. 

If the boilers are to be opened fires are allowed to die out or 
are hauled immediately. If time permits the former plan is pref- 
erable, as the change in the condition of the boiler is more 
gradual. When the fires are finally hauled and the furnaces, back 
connections and tubes cleaned out, the ashes, soot and clinker 
are wet down and piled away until they can be disposed of to 


the ash barge, as few harbor regulations allow the dumping of 
ashes overboard. In wetting down the fires after they are hauled 
out on the fire room floor, or in wetting down ashes at any 
time, care should be taken not to wet the fronts of the boilers 
or the mouths of the ash pits. The local chilling will not im- 
prove the quality of the steel, and the alternate wetting and dry- 
ing will increase the opportunities for surface corrosion. For the 
same reason damp ashes should never be piled up in contact with 
the boiler or furnace plates, as in many instances serious cor- 
rosion has resulted from a neglect of this precaution. 

The fires being burned out or hauled, some engineers pro- 
ceed to blow the boilers down immediately. This plan, however, 
cannot be recommended and should not be adopted unless the 
time available for examination and repairs is so short as to make 
it absolutely necessary. It is far better to let the steam condense 
and the water gradually cool, and then draw it out by means of 
a pump, or in some cases run it into the bilge. In this way the 
boiler cools more gradually and the structure is left in better 
condition, while on the water side the scale and incrustation will 
usually be made softer and more easily removed than when the 
boiler is blown down with steam on. 

[3] Emergencies and Casualties. 

(i) Foaming and Priming. These terms refer to a disturbed 
condition of the water in the boiler, of such a nature that the 
water level is more or less uncertain in location, and the steam 
space is partially filled with foam or a mixture of foam and water. 
In severe cases of foaming, steam seems to be given off from 
almost the entire mass of water in the boiler, causing it to rise 
bodily as foam and water and fill the whole steam and water 
space, thence entering the steam pipe and passing on to the 
engine. In other cases the water seems occasionally to rise in 
gulps, nearly unmixed with steam, and entering the steam pipe 
pass on to the engine. The terms foaming and priming are often 
used as meaning practically the same thing. Where a difference 
is implied, foaming is understood to apply more especially to 
the uplifting of the mixed steam and water as foam, while prim- 
ing may refer more particularly to the lifting of water as such, 
and its passage over into the engine. There is, however, no 
clear line of distinction between the two kinds of disturbances, 
and there are all grades intermediate between the extremes. 


Foaming may be due to the presence of certain forms of oil 
or grease, or to the excess of soda used for scale prevention, or 
to other impurities in the water, or to the demand for steam 
too large in proportion to the steam space in the boilers. A sud- 
den change in the character of the feed water may also produce 
foaming. In former days when jet condensers were in common 
use, boilers were liable to foam in passing from sea water to 
fresh water, especially if the latter was muddy, and again in 
passing from fresh water back to sea water. In modern prac- 
tice foaming is due either to the presence of oil, or to the extreme 
demand for steam from the boiler. In the former case the oil 
must be removed from the boiler by a free use of the surface 
blow, and kept out by a proper filter. In the latter case the 
engine must be slowed and the demand for steam reduced to an 
amount which the boiler can supply without the danger of such 

As a result of foaming the engine slows down, power and 
speed are lost, while due to the possible inability of the relief 
valves to handle all of the water coming into the cylinders, there 
may be serious danger of breakdown. There is also danger to 
the boiler in foaming, because the water level cannot be known 
with certainty, and plates or tubes may become overheated, with 
danger of collapse and rupture. The tendency to foam is, there- 
fore, a symptom of serious import, and no steps should be neg- 
lected to discover and, if possible, to remove the cause. 

(2) Feed, Pump. In modern practice an auxiliary feed-pump 
is always provided except perhaps in very small craft. If then 
the main feed-pump refuses to work, the auxiliary pump must 
be brought -into use while the other is under examination. 

The chief causes which may disturb the operation of a feed- 
pump are the following : 

(a) Jamming of check valve or other closure in the de- 
livery pipe. 

(b) Water in the steam cylinder. 

(c) Derangement or sticking of the steam valves. 

(d) Jamming or sticking of the water or steam plunger in 
its cylinder. 

In addition to these causes which may affect or prevent the 
motion of the pump, the following causes may prevent it from 
throwing water into the boiler, even though its movement may 
be entirely regular : 


(e) Split in the feed-pipe, or valve open, allowing the escape 
of the water at some unexpected point. 

(f) Excessive wear of water plunger. 

(g) Split or leak admitting air on the suction side or into 
the suction pipe. 

(h) An excessively high temperature of the feed water. 

(i) Derangement of the suction or discharge valves. 

To make sure that the delivery pipe is free, one or more 
feed-checks and the air-cock on the pump may be opened. If 
there is no movement of the pump or no discharge from the 
cock it may be concluded that the trouble is located elsewhere. 

The drain valves in the steam cylinder should then be blown 
out freely, and if there is still no inclination to start, the trouble 
is presumably with the valve gear. 

A fresh supply of oil should be admitted to the valve-chest, 
and an attempt may be made to work the tappets or other 
valve mechanism by hand. In many cases this will suffice to 
start the pump off at its regular gait. If it does not, the chances 
are that the trouble is more serious, involving the stopping up 
or clogging of some of the auxiliary ports or passages, or the 
sticking, jamming or excessive wear of some part of the valve- 
gear. A removal of the bonnets and complete overhaul can 
alone lead to a discovery of the difficulty in such cases. 

The jamming or rusting of the steam or water plungers in 
their cylinders could only result from long disuse and gross 
neglect, and can hardly be considered as of likely occurrence in 
routine work. 

If the feed-pump is working properly and throwing water 
into the boiler, the chamber of the check-valve will be relatively 
cool, there will be a click as the valve rises and falls, the air cock 
at the pump will show a stream, and the water will rise in the 
boiler gauge glass. At the same time an experienced eye and 
ear will detect by the manner of the pump, by the way it moves 
and by the character of the sounds, whether or not it is throwing 
water. If then the pump works, but does not seem to be throw- 
ing water, we must have resources to causes such as those men- 
tioned in (e)-(i). 

There are here two chief questions to be answered. First, is 
the pump getting water? and second, if it is, where is it going? 
The air cock will usually serve to answer the first question. If 
water appears here, and if the pump shows by its action that it 


is handling water, it is evident that there must be escape at some 
unexpected point. The feed-pipe must then be carefully ex- 
amined for leaks, and all valves or connections leading to or from 
it should be examined to make sure that the water is not escap- 
ing in some such way. In one case coming under the author's 
notice the main feed-pipe was fitted with a small branch leading 
to the forward tank. This branch was closed off from the feed- 
pipe by a globe valve. Due to the ignorance or carelessness of 
some attendant, this valve was jammed wide open instead of 
being jammed hard shut. At low or moderate pressure the feed- 
pump would throw enough water to feed the boiler in spite of 
this leak. The trouble was therefore not discovered until a full 
power run being started, the demand for water was greater and 
the leakage as well, so that the boiler was soon short of water, 
and the run was lost. 

If no trouble is found in the feed-pipe, the difficulty may be 
sought in a very loosely fitting or badly worn water plunger. 
Such a plunger will discharge water into the air or even against 
a low boiler pressure, but may not be able to force it against the 
regular pressure in the boiler. 

If from the evidence of the air cock and general behavior 
of the pump it is evident that no watef is being handled, 
the suction pipe and plunger rod packing should be examined 
for air leaks. 

Where there is some considerable lift from the hot well to 
the feed-pump, an unusually high temperature of feed-water, on 
account of the vapor formed, will sometimes prevent the pump 
from taking water. In good practice, of course, the hot-well is 
above, or at least not below the feed-pump, so that this difficulty 
is not likely to arise. If such should prove to be the trouble the 
feed-water must be cooled and the difficulty will be removed. 

If none of these causes seem to explain the failure to draw 
or discharge water, then the trouble is probably to be found in 
the suction or discharge valves, and the necessary bonnets or 
covers must be removed to allow of their examination. 

In any search for the cause of the trouble in the feed-pump, 
the details may, of course, be modified according to the circum- 
stances, and the above suggestions are more especially intended 
to illustrate the principle that in such a search the trouble has 
often to be found by a continued elimination of one thing after 
another, taking those which are most readily examined, and thus 


localizing the difficulty as quickly and as readily as possible. 

(3) Check-Valve Jammed. If the feed-pump seems to be in 
proper condition except that it slows down and stops when out- 
lets excepting a particular check-valve are closed, if furthermore 
the check-valve chamber is hot, there is no click, and the water 
does not rise in the glass, we may conclude that the check-valve 
is jammed on its seat. In former years this was sometimes due 
to unequal expansion of the valve and seat, and nipping of the 
former by the latter. In modern practice with good design and 
an angle of valve seat not steeper than 45 degrees, such an oc- 
currence is very rare. The valve may also become jammed by 
the bending or other derangement of the stem or wing guides, 
or by the lodging of some foreign body within the chamber. If the 
trouble is due simply to unequal expansion the usual treatment is 
to wrap the valve chamber in waste or clothes, and then to pour 
on cool water, thus reducing the temperature. At the same time 
the chamber may be tapped near the valve seat with a hammer 
or bar. In any ordinary case of sticking due to unequal expan- 
sion the result will be to free the valve from its seat. If this does 
not avail, then the stop-valve between the check-valve and 
boiler must be closed and the check-valve cover removed, so as 
to allow an examinaton of the interior. In good, modern prac- 
tice at least two check-valves, main and auxiliary, are provided 
for each boiler, so that there should be no danger of low water. 
If, however, only one check is provided, or if there is any ques- 
tion of shortness of water while the necessary repairs are being 
made, the stop-valve should be closed to stop the draft of steam 
and the usual precautions taken when stopping suddenly. (See 
[2] above). 

(4) Bursting of Water Gauge Glass. This occurrence is by 
no means uncommon, and under ordinary circumstances is of 
relatively small importance. With the type of gauge glass fitting 
having self-closing valves, as described in Sec. 17 [8], the flow 
of water and steam is automatically shut off and the new glass 
is readily put in. Without such provision the shut-off valves 
must be shielded from the discharge in the manner most readily 
effective, and then quickly closed. In setting a new gauge glass 
care should be taken to see that the fittings are well lined up, 
so that when screwed down there will be no bending strain on 
the glass. If any pronounced strain is thus set up on the glass 
by the fitting, it will be almost sure to break in a short time. 


The packing rings should also be screwed down no tighter than 
barely sufficient to keep the joint tight. To make the closing 
of the supply valves as short a job as possible, in case it should 
become necessary, it may be recommended to open them only 
enough to insure a good supply of steam and water to the glass, 
rather than wide open. Two turns of the valve will be usually 
sufficient, and it is then much more quickly closed than when 
open five or six turns. 

(5) Low Water. The occurrence of low water or the ab- 
sence of water from the gauge glass is one of the most serious 
emergencies which can arise in the fire-room. Immediate ac- 
tion is called for, and the most serious consequences may result 
from a mistake, or indeed may result in spite of whatever may 
be done. 

If there is reason to believe that the water has but just dis- 
appeared from the glass and the lower gauge cock gives indica- 
tions of water or of very moist steam, it may be fairly assumed 
that the level of the water is not below the tubes or combus- 
tion chamber tops, and in such case there will be no immediate 
danger of overheating or collapse, at least so far as the level of 
the water is concerned. The feed may therefore be put on 
strong without hesitation, and if the diagnosis has been correct 
the water will soon reappear in the glass and the incident is at 
an end. It may be considered prudent, however, to check at the 
same time the draft of steam from the boiler, and to deaden the 
fires to some extent by some of the methods referred to below. 
When the water reappears the boiler can then be put on its 
regular work as before. 

In the more serious case when the location of the water is 
quite unknown and the gauge cocks and glass give no indica- 
tions, there is some diversity of opinion as to the best procedure. 
The chief point of difference relates to the propriety of imme- 
diately putting on a strong feed. It has been claimed that if 
the plates were red hot, so much steam would be suddenly gen- 
erated as to rapidly increase the pressure, and burst the boiler. 
On the other hand, it has been pointed out that the amount of 
steam which can be thus formed is in reality comparatively 
small, and that its formation cannot be instantaneous, nor even 
especially rapid, since its formation will extend over the period 
while the water is rising over the heated surfaces. In no way 
then could any great amount of steam be generated with es- 


pecial rapidity, and it is hard to see how its formation could 
take place more rapidly than would be provided for by the 
natural outflow to the engine, and by the safety-valve, if need 
arose. To obtain some definite information on these points, ex- 
periments were carried out by a Steam Boiler Insurance Com- 
pany some few years ago, in which the plan of putting in cold feed 
upon overheated plates was followed. The boiler could not be 
burst by the operation, and no very pronounced elevation of 
pressure was produced. So far as the results of these experi- 
ments went, it would seem to be safe to put on the feed imme- 
diately in such an emergency as we are now discussing. This 
conclusion seems also to be borne out by the results of such 
practical experience as is obtainable. There are, however, dif- 
ferences of opinion on this point, and while the author would 
follow this plan in such a case, it should not be denied that 
many good authorities would consider it unwise, at least in the 
earliest stages of the measures, to be taken. 

Aside from getting the feed into the boiler as soon as 
prudence will permit, the other great point is to deaden the fire. 
If the heating surfaces have not collapsed when the condition is 
discovered, it is hardly likely that the plates can be at more than 
a very dull red, and if the supply of heat can be effectually 
checked, further trouble may be averted. 

To this end a plan often followed, especially in former 
years, was to haul the fires. This, however, seems very unwise 
indeed. While being hauled the fires will burn up all the more 
fiercely, and for a few moments the heat supply will be in- 
creased rather than decreased. Instead of hauling directly, 
some engineers prefer to have the fires dumped into the ash-pits 
by dislodging a few grate-bars, and then to haul from there. 
This plan seems but little better than the other, and in any 
event the immediate result will be to increase the amount of heat 
given off at the very time when it should be decreased. 

A far better plan seems to be to deaden the fire with either 
moist ashes or coal. If a pile of wet ashes is at hand they should 
be thrown immediately on the fire, and will be found a most 
effective means of deadening the burning coals. In default of 
wet ashes, wet or even dry coal may be thrown on and the fir$ 
simply smothered. At the same time the dampers should be put 
up and furnace doors left open, thus checking all draft and 
stopping the formation of heat. 


As between these two operations, the deadening" of the fire 
and the getting in of feed-water, the author is of the opinion 
that the deadening of the fire is of the more immediate impor- 
tance, and should be attended to first, because it will produce 
the most immediate effect over the whole heating surface, and 
will serve most effectually to rapidly check the further heating 
of the plates. Putting in the feed water will then complete the 
cooling, but the direct operation is slower, and more local in its 
influence, and is therefore of relatively less importance. 

In all such emergencies much, of course, will depend on 
the special circumstances, but if the two principles be kept in 
view that the fire must be deadened and the water restored, the 
details may be left to good engineering judgment to execute. 

After the water has again appeared in the glass, and if no 
ill effects seem to have resulted to the boiler, the fire may be 
again gotten into condition and the boiler put on its regular 
routine. If, however, there be any doubt whatever regarding 
the possible results to the boiler, no chances should be taken, 
but the boiler should be disconnected from the others, the safety- 
valve raised and the steam blown down, while the fires should 
be hauled or allowed to die out and the boiler allowed to 
cool down. It should then be carefully examined for symp- 
toms of distress or collapse, and if any such are found they must 
receive proper attention before the boiler is again set to work. 

(6) Collapse of Furnace Crounis or Combustion Chamber 
Plates. The collapse of a part of the heating surface is due either 
to an overheating of the plates or tubes, or to a gross error in 
the design. We may dismiss the latter as not liable to occur in 
good practice. Overheating may result from either low water, 
or from a coating of scale or of oil and scale on the water side. 
In the former case with no water to absorb the heat the natural 
result is an overheating of the plate until it becomes red hot, 
followed by its collapse and rupture. When the overheating is 
due to the presence of a coating of scale, the gradual bulge or 
start toward a collapse may result in cracking off the coating 
and in letting in the water to the plate. This will cool the 
metal, restore its strength, and thus put an end to the operation. 
In some cases when the covering is a mixture of scale and 
oil, the overheating of the plate will result in burning off or 
volatilizing the oil, leaving the scale as a fine powdery deposit. 
This readily admits the water to the plate, and thus the metal 


may be cooled and restored in strength, and further bulging 
prevented. This nice adjustment of heating, burning or crack- 
ing off, cooling and re-strengthening before final rupture, does 
not, however, always occur. Before the re-cooling is effected 
rupture only too often results, and the contents of the boiler 
are more or less completely emptied into the fire-room, with 
consequences always severe and sometimes fatal. 

If it is discovered that the furnace crowns have come down, 
but without final rupture, or if any portion of the heating sur- 
face has suffered collapse or bulging, but without final rupture, 
it may be assumed that the overheating was due to scale or oil, 
and that the change of form or the overheating has resulted in 
getting rid of the coating, and in readmitting the water to the 
plate. So that if rupture has not yet occurred, it is probable 
that the plate is safe for the time being, and in such case the 
fire may be first deadened and then hauled, the boiler shut off 
from the others and allowed to cool down, and then examined 
as to the nature and extent of the injury sustained. 

(7) Collapse and Rupture of Furnace Crowns or Combustion 
Chamber Plates. In the event of rupture following upon col- 
lapse, the chief thought after the safety of human life must be 
for the remaining boilers while the fire-room is in such condi- 
tion that it cannot be entered. 

The general nature of the steps which may be taken in such 
case are discussed in the following section, and while naturally 
judgment must be depended on for many details, it is readily 
seen that the two main points are as follows : 

(1) To isolate the injured boiler. 

(2) To safeguard the remaining boilers from injury due to 
low water. 

(8) Serious Leakage in Boiler Tubes. A serious leak may 
suddenly develop in one or more of the tubes. A split in the 
metal, a collapse due to overheating, or perforation due to deep 
pitting or general corrosion may give rise to such an occurrence. 
If the hole or holes are not too large the immediate conse- 
quence will not extend beyond a more or less complete filling of 
the fire side of the boiler with water and steam, and a more or 
less pronounced checking or deadening of the fire. In such 
case the feed should be looked after to see that the water does 
not get low in the boiler, while preparations are made for plug- 
ging the tubes. For this purpose a tube stopper or plug is used, 


of which there are many varieties. A standard type of stopper 
consists of two heads or tapered plugs which make a joint 
within or against the ends of the tube, and are held in place by a 
rod running through the tube, threaded at the ends and pro- 
vided with nuts for holding them up to their position against 
the ends of the tube. To fit this stopper it is necessary 
to enter the combustion chamber to adjust the back end, but 
once properly fitted, it may be depended upon to fulfil its pur- 
pose. There are also special forms of tube stoppers which may 
be inserted and adjusted from the front end only. It is more 
difficult to make a tight joint with the latter than with the former 
stopper, but they can be fitted without drawing the fire, and 
therefore in an emergency may prove of great value. Plugs of 
soft pine are also used for temporary purposes. These are 
pushed in from the front until they cover the leak, and the ex- 
pansion due to soaking with hot water is depended on to make 
a tight joint. See also Sec. 42 [7]. 

If, however, the holes in the tubes are of considerable size, 
so great a quantity of water and steam may be liberated as to 
make it impossible to remain in the fire-room. In such case 
the fire will usually be put out as well, or at least so deadened 
that no further danger of collapse due to overheating need be 
feared, even should th> water become low in the boiler. The 
general safety of the boiler itself is thus secured, but the other 
boilers, connected through the main steam pipe, will continue 
pouring out their steam through this leak, and as long as this 
condition continues it will thus remain impossible to return to 
the fire-room to attend to the other boilers. The first care after 
escaping from the fire-room, or before effecting escape, if pos- 
sible, should be to close the stop valve which connects the boiler 
to the others, and thus to localize the trouble to the one boiler. 
If the stop valves are arranged so as to be worked from the 
deck above, as is frequent in good modern practice, this may 
be readily accomplished. If at the same time the safety valve 
on the injured boiler can be opened, the pressure will soon be 
blown down, and with good ventilation from the fire-room it 
will be possible in a short time to re-enter it and give the needed 
attention to the other boilers. In the meantime also, the en- 
gines may be slowed down so as to reduce the demand for 

Where the feed-pump is located in the engine room, it will 


be possible, if the checks have been left open on the other 
boilers and closed on the injured boiler, to feed by judgment at 
a rate which will keep these boilers safe from all danger of low 
water. The object of these various steps is, of course, to isolate 
the injured boiler and safeguard it from anything more serious, 
and to provide for the safety of those remaining ; and while cir- 
cumstances may alter the details of the measures which should 
be taken, the above suggestions will serve to illustrate the main 
points to be looked after. 

When return to the fire-room is possible, and after the other 
boilers have received the necessary care, attention may be given 
to the injured boiler; the fires may be hauled, and after cooling 
down, the nature and extent of the damage investigated. 

(9) Rupture of Steam Pipe. In the case of a ruptured steam 
pipe the valves controlling the flow of steam to the point of 
rupture each way should be closed at the earliest possible 
moment, as the escape of steam will soon make it impossible to 
remain in the fire-room with safety. If it is arranged to work 
the stop-valves from the deck above, this is readily effected, 
and the trouble thus gotten under control. Self-closing valves 
are also often provided in modern practice, as referred to in 
Section 17 [3]. It is not possible, however, to so fit these that 
they will always act, or that they will shut off the steam in more 
than one direction. In one way or another, however, the first 
attempt must be to shut off the escape of steam. The next 
thought must be for the boilers, to safeguard those which may 
have been shut off from the engine from danger due to increase 
of pressure, and those which still remain connected to the engine 
from danger due to low water. The particular steps suitable 
for attaining these objects have been already sufficiently dis- 
cussed, and no further mention will be here necessary. 

(10) Casualties With Water -Tube Boilers. With water-tube 
boilers the same general emergencies are liable to arise as with 
fire-tube boilers, and may be met in the same general way. It 
should be remembered, however, that, due to the small amount 
of water carried, the results due to shortness of water will come 
much more rapidly than with fire-tube boilers, and promptness 
in action is all the more necessary. In such boilers the tubes 
are most liable to suffer through shortness of water, and any 
considerable; overheating is likely to result in their rupture. 
With the small tube type the most serious results of such an 


accident are usually confined to the emptying of the contents of 
the boiler into the fire, and the effectual deadening or extinction 
of the latter. Here, as before, however, with more than one 
boiler the trouble must be localized by shutting off the boiler 
with the ruptured tubes from connection with the others. In 
other cases, however, with large tube types especially, or with 
the rupture of steam or water drums, the consequences may be 
more serious, resulting in driving every one from the fire-room, 
or even in loss of life. The same general principles apply here, 
however, as with fire-tube boilers, and good judgment must be 
depended on for the details suitable to the occasion. 

[i] Getting Under Way. 

In the engine room the same as in the fire-room, a general 
inspection is first in order, more or less detailed and extensive, 
according to the time the machinery has been out of use, and 
the degree of acquaintance with its various features and pecu- 
liarities. If the engine has been laid up for any length of time 
a detailed examination will of course be required similar to that 
referred to at a later point in Sec. 43. We here assume, how- 
ever, that no such general overhauling is needed, and that the 
machinery is to be supposed in a working condition. 

A good general idea should first be obtained of the lead 
of the principal piping systems as noted in Sec. 25, and especially 
of the main and auxiliary steam and feed systems. These lines 
of piping should be looked over, the location of the valves 
noted, and where permissible the valves should be opened or 
closed to insure their being in working order, and then left in 
the condition desired for getting up steam. 

The various parts of the main engine will be looked over 
so as to insure, so far as an external examination can, that 
everything is in proper working order. 

The various auxiliaries will be looked over in the same 
way, and the results of this general examination being satis- 
factory, steps may be taken to test the various parts of the ma- 
chinery under steam as soon as dt is ready. 

We have already in Sec. 38 pointed out the importance of 
trying the feed-pumps and getting them into working order as 
soon as possible, in order to insure the proper supply of feed- 
water to the boiler. Next in order may come the circulating 


pump, the engine of which is started at moderate speed, and the 
main injection and discharge valves opened. In starting all of 
this auxiliary machinery, proper precautions must, of course, 
be observed in regard to freeing the steam cylinders of con- 
densed water by means of the relief valves, as noted in Sec. 38 [3] 
in connection with the feed-pump. The circulating pump being 
usually below the level of the water outside the ship, it naturally 
floods itself so that no trouble should be met with in getting it 
to take water. Assuming the air-pump independent, this may 
be started next and put at a moderate pace, or sufficient to main- 
tain a vacuum of 15 to 20 inches. 

The electric light engines, if not in operation from the 
donkey boiler, will also be looked after and started in due time, 
as well as any other auxiliary machinery whose operation may 
be required for getting under way. 

In case the main engine since the last time used has been 
subjected to any adjustment or overhauling, it will be well to 
turn it completely over with the turning engine once or twice 
in order to make sure that everything is clear and in running 

In the meantime, while the auxiliaries are being gotten into 
operation, the steam will have been admitted to the jackets, if 
there are any, and to the cylinders through the main stop and 
throttle valves. Here, as noted in Sec. 38 [i], the object in view 
is to avoid any sudden change in the temperature or heat condi- 
tion of the machinery. A good method of gradually warming 1 
up the engine is to just unseat the main stop and throttle valves, 
and then with the links in the ahead gear, say, to slowly turn 
the engine over ahead with the turning engine. This will allow 
the steam to work its way through the engine, warming up the 
entire series of cylinders and bringing them practically to their 
working temperatures. The water condensed must, of course, 
be allowed to escape by opening the relief and drain valves. 
During this period the reversing gear will also be warmed up 
and tried under steam until it works properly and throws the 
links smoothly from one side to the other. 

When the engine has thus become well warmed up, the 
turning gear will be disconnected and locked out of gear, and im- 
mediate preparations made for turning over under steam. At 
this point the question of lubrication must be borne in mind, and 
while the regular schedule of oiling, etc., need not be started 


until the ship is fairly under way, still a moderate provision of 
oil may be made to the more important bearings, and if the en- 
gine has been out of use some little time, it will be well to work 
oil into the principal bearings during the preceding operation 
of the main engine with the turning gear as suggested above. 

Before turning over under steam the deck officer should 
be notified in order that the hawsers securing the ship to the 
dock may be looked to if necessary, or the presence of anything 
about the stern which might foul or jam the propeller may be re- 
ported back to the engine room. Everything being in readiness, 
the main stop valve is opened slowly to full opening, and steam 
is turned on the reversing gear. Then the main throttle being 
still closed the links are thrown back and forth a few times, the 
passover or starting valves opened, and the throttle opened 
moderately. If the engine does not start off in one direction 
the links are thrown into the other gear, and if everything is in 
the proper condition the engine will start in one direction or the 
other after a few see-saws of the links. The hand relief valves 
are, of course, operated at the same time in order to aid in 
freeing the cylinders of any water which may collect there, or 
which may enter with the steam. Often the engine will move 
a little way, but the high pressure, or one of the other pistons 
will not pass the center. This is on account of the water in the 
cylinders, and is especially liable to occur if the engine has not 
been well warmed up, or if the steam pipe has not been properly 
drained. In such case the water must be worked out through 
the relief and drain valves, the links in the meantime being 
moved back and forth. In answer to this the piston will see- 
saw up and down, getting gradually nearer the center, and 
finally when the water is sufficiently cleared out, passing over 
and continuing the revolution. As the main engine is thus 
started the circulating pump and air pump, if independent, will be 
started up at the increased pace suitable to the amount of steam 
passing through the engine and into the condenser. After thus 
running for a few minutes, or until everything seems to be in 
proper running order, the engine is stopped, and the signal for 
the regular start is awaited. The object of thus turning over 
under steam is simply to make sure that everything is free and in 
working condition. Little of course, can be told regarding the 
adjustment of the various bearings, etc., or their liability to heat 
or pound. If the machinery is new or has undergone any con- 


siderable readjustment, or has been out of use a long time, it 
should have a dock trial of some considerable time, in order to 
determine the various points, and to bring out any defects liable 
to present themselves in the course of a continuous run. 

Naturally the time when the ship is to start will be known 
to the engineer, and these various preparations will be so timed 
that soon after the final turning over under steam, the signal for 
the regular start may be expected. 


In connection with the operation of a marine engine it will 
be instructive to note in order the names of the various parts 
through which the steam passes from the boiler until it returns 
again as feed water to its starting point. Starting, then, with 
its formation in the boiler, we have the following route for the 
case of an ordinary triple-expansion engine : 

Dry-pipe Safety valve chamber Boiler stop-valve 
Boiler steam-pipe Main steam-pipe Main stop-valve Main 
throttle-valve High pressure valve-chest Steam ports and 
passages High pressure cylinders Steam passages and ports 
Exhaust side of valve Exhaust passage Exhaust pipe to in- 
termediate valve-chest and cylinder as above for the high pres- 
sure Exhaust pipe to low pressure valve-chest and cylinder 
as above for the high pressure Exhaust pipe to condenser 
Condenser Air-pump suction Foot-valves Air-pump cham- 
ber Bucket-valves Delivery-valves Hot-well Feed pump 
suction Induction valves Feed-pump barrel Discharge 
valves Feed-pipe Check-valve Boiler. 

In addition a separator may appear between the boiler and 
the engine, and the filter and feed water heater between the hot- 
well and the boiler. 

[a] Routine Operation. 

In the routine operation of the main engine and of the 
other machinery in the engine room, the following are the points 
requiring chief consideration : 

(a) The proper provision of oil and other lubricant in suit- 
able quantities and at proper intervals, or continuously, accord- 
ing to the nature of the oiling gear in use. 

(b) A constant watch over the general conditions of 
operation of the machinery in order that any symptom or 


sign of derangement or disturbance may be noted, and the 
proper steps taken for its control or removal. 

The chief points relating to lubrication have been already 
discussed in Sec. 24 [12]. The watch over the general condi- 
tions extends to all features and depends to such an extent upon 
the special circumstances that only a few general hints can be 

First, regarding the sounds which accompany the opera- 
tion of the machinery and the part the ear may take in de- 
tecting symptoms of disturbance. The operation of the main 
engine and of the various parts of the machinery individually is 
accompanied by more or less plainly marked sounds or noise or 
combination of sounds. These in the end tend to combine them- 
selves into a kind of resultant rumble, click, and rattle, which 
often remains quite constant in character, and so comes to have 
a kind of individuality of its own. To a person accustomed to 
the regular sounds of the engine room, the ear is often a deli- 
cate means of detecting any departure from the regular routine, 
and often the first indication of some disturbance will be fur- 
nished by a change in the character of the sounds produced. 
In particular, any unusual pound, jar, squeak or rattle should be 
located as soon as possible, and its cause investigated. 

In some cases assistance in the detection or location of a 
pound or knock may be gained by the use of a convenient piece 
of metal, such as a spanner, or bit of pipe, one end being placed 
to the ear and the other against the cylinder, valve-chest, or 
other point nearest to where trouble is expected. At the same 
time too much reliance must not be placed on the ear to the 
neglect of other means of observation. In fact in the modern 
engine room all of the available senses keenly on the alert will 
be found none too many for the proper watch and care over 
the machinery in use. 

The danger of heating, due to insufficient lubrication, poor 
adjustment or bad condition of bearing, is one which the ear 
will often aid in detecting, but the chief reliance must be placed 
on the sense of feeling and on the nose. With the necessary 
skill most of the important bearings may be felt by the hand. 
Caution must be used, however, so that the hand may not be 
caught or jammed. This part of an engineer's training is one 
which can be learned only by observation and cautious trial. If 
the heating of the bearing passes beyond a moderate elevation 


of temperature, the oil will become correspondingly heated and 
will give off a burnt odor, or perhaps will smoke freely, thus show- 
ing plainly the existence of trouble. The nose and eye will thus 
come in as factors in detecting trouble of this character. 

Small steam leaks at joints, stuffing-boxes, etc., will make 
themselves plainly visible, and should receive such treatment as 
the circumstances may require, in order that they may be 
closed up. It must not be forgotten that every steam leak 
means a loss of both heat and fresh water. 

The vacuum in the condenser will depend not only on the 
proper operation of the air-pump, but also on the reduction of 
all possible air leaks which might admit air to the low pressure 
cylinder during the exhaust, or to the steam side of the conden- 
ser. All such stuffing-boxes, joints, etc., must therefore receive 
careful attention, especially if the vacuum is not what it 
should be. 

Water coming over into the cylinders from the boilers pro- 
duces a crackling or snapping noise, which is readily recog- 
nized. The automatic relief valves may, of course, be depended 
on, or the hand gear may be operated to aid in removing the 
disturbing cause. 

In stopping momentarily the throttle is closed and the links 
are run to mid-gear, no other change being made, and every- 
thing remaining ready to start again at an instant's notice. In 
stopping for a known period of time of any considerable dura- 
tion, means should be taken to stop the flow of oil to the bear- 
ings by the closure of the feeding valves or the withdrawal of 
wicks, according to the means in use. If the stop is only tem- 
porary, and the engines are to be kept in readiness for starting 
again, the further steps taken will be only such as will serve 
to bring the machinery into its condition just previous to getting 
under way. That is, the circulating and air-pumps, so far as in- 
dependent, may be slowed somewhat, while by the aid of the 
steam jackets, if fitted, and steam which is allowed to flow past 
the stop and throttle valves, the main engine is kept warmed up 
and ready for operation at short notice. 

If the stop is to be of longer duration and the steam is to 
be shut off the engine, the stop and throttle valves will be 
closed, the air and circulating pumps shut down, and steam 
shut off the reversing engine, jackets, etc. The various drain 
valves and drips will be left open so as to free the machinery, 


as far as possible, of all the water formed by the gradual con- 
densation of the steam. At the same time the flow of oil to the 
bearings will be shut off and such measures taken with respect 
to the oiling gear as the circumstances may require. 

[3] Minor Emergencies and Troubles. 

We will now consider briefly the steps to be taken in the 
event of the more commonly occurring troubles, some of which 
have been mentioned in the preceding paragraph. 

(i) Derangement in the Oiling Gear. In sight-feed apparatus 
this is readily detected, and without loss of time the trouble 
must be located and remedied, the oil in the meantime 
being supplied to the bearing in question by hand. The trouble 
in such cases usually arises from a clogging up of some of the 
pipes or passages, and as noted in Sec. 24 [12] all such pipes 
should be put up with union joints so that they may be readily 
taken down, cleaned and replaced. 

(2) Hot Bearing. This is one of the most important of the 
minor troubles which may arise in the engine room, and one 
which may lead to serious consequences in case the proper steps 
for its control are not taken in time. A hot bearing may arise 
from a variety of causes, among which the following are the 
more important. 

(a) Lack of lubrication. 

(b) Lubricant too thin so that it will not remain in place in 
the bearing and sustain the load. 

(c) Improper adjustment, the amount of clearance between 
journal and brass being too small. 

(d) Lack of alignment in the machinery, as a result of 
which the bearing is excessively severe on certain parts, thus 
forcing out the lubricant and causing the surface to nip and 

(e) Bearing surface not of sufficient area to carry the load 
or take the work put upon it without an undue rise in tem- 
perature. This means, of course, either that the design is faulty 
or that the machinery is worked beyond the loads for which it 
was intended. 

(f) Bearing surfaces rough and uneven, due to the poor 
workmanship, or as a result of serious heating on a previous 

If the trouble is due to a lack of lubrication simply, and 


is discovered in time, an abundant supply of oil will be usually 
sufficient to control the condition and to gradually bring the 
bearing back to its normal temperature. If, however, the tem- 
perature rises considerably, the journal may expand more than 
the bearing brasses, so that the clearance will be decreased and 
the brasses will pinch the journal, thus introducing a further 
source of trouble as noted in (c) above. If this is not soon 
relieved, the metal surfaces will nip and the softer of the two 
will begin to abrade or "cut." This is always the bearing 
metal, and the resulting condition, in consequence of which the 
smoothness of the surface is destroyed, will tend simply to make 
matters still worse, to generate more heat, expand the parts 
still more, perhaps nip the surfaces still more tightly, and so cut 
the worse, until the bearing metal melts and runs out. 

The treatment of a heated bearing involves two chief items, 
viz., the removal of the cause and the restoration of the bearing 
to its normal condition. 

We may remove the cause entirely, of course, by stopping 
the engine, and in an advanced case of trouble, such as just 
described, this may be necessary. Otherwise we may reduce the 
cause by slowing down somewhat, and thus decreasing the 
amount of work thrown on the bearing. We may further de- 
crease the cause by easing up the bearing cap and thus increas- 
ing the clearance between journal and bearing surface. This, 
however, can only be done to a slight extent, else trouble will 
be met with from too great clearance and the consequent pound- 
ing in the joint. 

A plentiful supply of oil, or other lubricant, will also aid in 
decreasing the cause and in restoring matters to their proper 

A decrease in temperature will also usually aid in removing 
the cause, and is, furthermore, of course, one of the chief steps 
in bringing the bearing back to its normal condition. 

To this end in the extreme case, it may be considered neces- 
sary to turn a stream of water on the bearing, thus to absorb 
and carry away the heat, and in many cases full power trials are 
run with streams of water playing for a considerable part of the 
time on various parts of the machinery in order to carry off the 
heat and so control the temperature. Water, however, is doubt- 
less used far more than is absolutely necessary, and far more 
than good engineering would authorize. If sprayed or run from 


a hose on the bearings it is almost certain to find its way in and 
an to the bearing surfaces, where it will prevent action of the 
lubricant. For this reason its use once begun, it may be neces- 
sary to continue, simply because the lubricant cannot lubricate 
in the presence of water. In the best modern practice, as indi- 
cated in Sec. 21 [n], provision is made for circulating water 
through hollow bearing blocks, and thus in the most effective 
way the water is able to remove the heat generated without 
coming into contact with the bearing surface itself. 

In case the machinery is not properly lined up, or the bear- 
ings are of insufficient area, or not in proper condition, only 
temporary relief can be looked for from the various means sug- 
gested above, the most effective of which will presumably be the 
operation of the machinery at a low or moderate power until 
such time as the needed readjustments, changes or repairs can 
be effected. 

To sum up the treatment for a hot bearing, the measures 
taken may be selected according to judgment and the special 
circumstances from the following: 


Easing up bearing caps. 

Slowing down and consequent reduction of load. 

Application of water. 

(3) Pounding. This condition may arise from several 
causes, chief among which are the following : 

(a) Bearings not in proper adjustment, too much clearance 
being allowed between journal and bearing metal. (See Sec. 
24 [12].) 

(b) Lubricant too thin and thus unable to retain its place 
in the bearing. 

(c) Valve events not properly adjusted, especially the ex- 
haust closure and following compression. 

Furthermore, an engine will often show at different speeds 
a marked difference in this respect, such difference being chiefly 
due to the increasing effect of the inertia forces with increase in 
the revolutions. 

If the trouble arises from the nature of the lubricant in 
use a change to a heavier oil may show an improvement. If, 
however, as is more commonly the case, it is due to faulty ad- 
justment, either of bearing or valve gear, or both, but little can 
be done while the engine is in operation, and the first oppor- 


tunity for overhauling and readjustment must be taken for a 
study of the conditions, both as regards the bearings them- 
selves, and the possibility of improvement by an adjustment of 
the compression. If the pounding becomes very severe," it may 
become necessary to slow down the engine and operate under 
less than the regular or full power until the proper examination 
and readjustment can be made. 

(4) Priming or Lifting Water. This emergency has been 
more particularly referred to in Sec. 38 [3] . In small quantities 
water produces a crackling or snapping sound in the cylinders, 
and the automatic relief valves may be allowed to take care of 
the situation, of if desired, the hand reliefs may be operated as 
well. If, however, the water comes over in large quantities the 
engine will slow down and work with an irregular and labored 
motion, which may be readily recognized as denoting this con- 
dition. In such case the throttle or main stop valve should be 
partially closed and the water gotten rid of as quickly as possible 
by the use of the relief valves. The engine will then operate at the 
reduced speed permitted by the partially closed valve, pre- 
sumably without further trouble. If on opening out again the 
tendency to lift water at full or ordinary power is persistent, the 
power must be reduced until the trouble is removed and the 
engine will operate continuously without disturbance of this 

(5) Vacuum Falls and Becomes Poor While the Condenser Be- 
comes Hot. Following are the chief causes which may lead to 
such a condition : 

(a) Insufficient condensing water from any cause. 

(b) Division plate in condenser head carried away so that 
water goes directly from inflow to outflow without going through 
the tubes. 

(c) Excessive inflow of steam caused by leakage either past 
low pressure piston or slide valve, or possibly in the ''bleeder" 
of "silent blow" if such is fitted. 

(6) Vacuum Falls and the Condenser Remains Cool. In such 
case the indications are that this condition is due to the presence 
of air not removed by the air pump, as may be caused by any 
one or a combination of any of the following : 

(a) Air-pump valves defective. 

(b) Leak in the condenser, either at the head joints or 
through a crack. 


(c) Soda or drain cocks open or leaking. 

(d) Low pressure piston rod stuffing box leaking air inward 
during exhaust stroke. 

(e) With "inside" piston valves on the low pressure cylin- 
der, leaky valve stem stuffing boxes. 

(f) Leak or obstruction in the pipe leading to the vacuum 


NO sooner has a boiler been completed than the various 
corrosive and destroying influences with which it is surrounded 
set to work on its destruction. We may conveniently consider 
corrosion as of two kinds, that due to oxygen and that due to 
an acid. These two are, however, by no means independent, 
and are often combined in very complex ways. The process by 
which oxygen combines with another substance is called oxida- 
tion, and the product of the operation an oxide. In the case of 
iron and steel the typical product is the ordinary red iron rust, 
or ferric oxide (Fe 2 O 8 ), consisting of about 56 parts -by weight 
of iron and 24 of oxygen. In order that oxidation or rusting of 
iron may continuously proceed at ordinary temperatures, how- 
ever, it is not enough that oxygen and iron shall be in contact. 
It requires the additional presence of moisture and carbon di- 
oxide (CO 2 ), small proportions of which are always present in 
the atmosphere. Oxygen and moisture alone act feebly and 
very slowly on iron, but when the four substances, iron, oxygen, 
moisture, and carbon dioxide, are all present together, the 
operation of rusting proceeds continuously and with vigor. 
Oxide is first formed, and this is reduced by the carbon dioxide 
to a carbonate, and this in turn breaks up, forming hydrated 
oxide (FeHO 2 ), setting free the carbon dioxide to continue the 
process. The hydrated oxide thus formed is furthermore elec- 
tro-chemically negative to iron, and thus helps on the operation 
as explained at a later point. If either the moisture or the car- 
bon dioxide is absent the oxygen will have little or no effect, 
and the iron will be protected. This is shown by the non-rust- 
ing of iron in perfectly dry air, even though there may be some 
carbon dioxide present ; or again, by its preservation in a weak 
alkaline liquid, as lime water, in which there can be no free car- 
bon dioxide. The piano wire used in certain forms of deep sea 
sounding apparatus, for example, is thus kept from corrosion 


under conditions which would naturally soon destroy its regu- 
larity and value for the purpose used. 

Acid corrosion means the attacking of a substance by an 
acid, the breaking up of the latter, and the formation of a new 
substance known as a salt, and composed of a part of the acid 
and of the substance attacked. Thus hydrochloric or muriatic 
acid (HC1), as it is commonly called, is sometimes present in 
boilers. This is composed of hydrogen and chlorine. When it is 
brought into the presence of iron or steel the chlorine leaves the 
acid, and joining with the iron, forms a salt known as ferrous 
chloride, or chloride of iron (FeCl 2 ). With iron rust and mu- 
riatic acid the result would be similar, the chlorine would join 
with the iron and form ferrous chloride, while the hydrogen of 
the acid would join with the oxygen of the oxide and form 

As before stated, acid corrosion and oxidation are very 
commonly both present, especially in the latter operation, and 
in fact the continued progress of oxidation with iron, moisture 
and carbon dioxide is dependent on the combined action of both 
operations. W T e shall not, however, deal further with the 
chemical details of corrosion in general, but proceed to a brief 
consideration of the causes, effects and remedies as related to 
corrosion in marine boilers. 

Taking first the exterior o'f boilers and of all exposed iron 
and steel work in general, it is clear that the conditions for con- 
tinued rusting are all present on board ship. The air is moist 
and there is likely to be present carbon dioxide in abundance. 
The only safe protection is, therefore, a covering which shall 
keep the air, moisture and carbon dioxide from contact with the 
iron. To this end metal paint or other equivalent coating is 
used wherever possible. Many small fittings, especially about 
the deck, are of galvanized iron, that is, iron covered with a thin 
coating of zinc. The latter metal is but slightly affected by the 
process of oxidation, and it, therefore, forms an efficient protec- 
tion for the iron. Brass, bronze and copper are also oxidized 
but slightly, and the oxide formed serves as a protective cover- 
ing to the metal underneath. For this reason, among others, 
many of the fittings about boilers and elsewhere are, as we have 
already seen, made of these metals. 

Passing in now to the fire side of the boiler, we find the ap- 
plication of paint or other protective coating impracticable. 


Here we must depend on the heat, which will so dry the air that 
it is no longer moist. That is, while water vapor may still be 
present in the air, there is so little compared with the amount 
the air could naturally contain at that temperature, that it is 
held by the air and is no longer free to enter as a factor into 
the operation of oxidation. Rusting in the usual way is, there- 
fore, very much retarded or prevented. To this fact we owe the 
general preservation of the furnaces, ash-pits, etc., from serious 
and continued corrosion. We here, however, run -into another 
danger in the extreme case when oxygen is present in excess, 
and both the oxygen and iron are very hot. The oxygen in 
such cases enters more readily into union with the iron, and if 
the temperatures should be sufficient, a different kind of oxide 
is formed, the black, or magnetic oxide (Fc 8 O 4 ), the same as the 
mill scale or forge scale, which forms when iron is worked at a 
red heat. The oxide thus formed may presumably be swept 
away by the scouring action of the draft, thus exposing a fresh 
surface to renewed attack. The back ends of the tubes seem 
especially liable to attack in this way, and particularly with hard 
forced draft. The cure for this trouble is found in the use of cast 
iron ferrules, as previously described. 

These ferrules protect the tube ends from the extremes of 
temperature, and also provide something for the hot oxygen 
to attack, while they are readily renewed. 

Turning now to the water side of the boiler, we find more 
serious trouble than with the fire side. There is likely to be 
more or less air in the feed water, either entering with the 
make-up feed, or occasionally drawn into the feed-pump and 
sent on to the boiler. There may also be free carbon dioxide 
liberated from the salts entering with the make up feed, and 
thus all the conditions for continuous rusting may be present. 
Even if free carbon dioxide is not present the formation of iron 
oxide, combined with electro-chemical reactions, as referred to 
later, may result in serious local corrosion. Furthermore, as 
the feed-water is heated the air is liberated, and the oxygen just 
at the instant of liberation seems to be especially active chemi- 
cally, and is thus all the more likely to attack exposed places 
than if allowed to remain in solution in the water, as at ordin- 
ary temperatures. 

Turning next to acid corrosion, mention may first be made 
of the serious trouble formerly experienced from the use of 


animal and vegetable oils for cylinder lubrication. Such an 
oil is a compound of a fatty acid and glycerine. When exposed 
to a high temperature the fatty acid and the glycerine become 
separated. If a substance such as soda or potash is present, 
the fatty acid combines with it and forms soap. This process is 
called saponification. If, however, no such substance is present 
the acid will be free to attack other substances as it may be able. 
Fatty acids attack iron feebly, but if long continued the result 
may be a serious corrosion, resulting in the formation of what is 
known as an iron soap. The temperature within the cylinders and 
boilers was quite sufficient to thus decompose the oil, and there 
would, under such circumstances, be set free in the boilers an 
amount of fatty acid depending on the amount of oil used in the 
cylinders and finding its way into the condenser and feed-water. 
There were thus present all the conditions necessary for the 
corrosion of the interior of boilers by fatty acids, and many seri- 
ous cases were laid, in part at least, to this cause. These trou- 
bles appeared especially with the introduction of the surface 
condenser, and the part which fatty acids might play being un- 
derstood, the use of animal and vegetable oils for the lubrication 
of the cylinders was abandoned, and in their place hydrocarbon 
or mineral oils are now used. Such oils are derived as one of 
the constituents of crude petroleum, and are not compounds of 
a fatty acid and glycerine. They are compounds of carbon and 
hydrogen, and belong to an entirely different class of chemical 
substances. They do not produce a fatty acid on being heated, 
and cannot, at least directly, take part in the process of boiler 

In modern practice, therefore, nothing but the best hydro- 
carbon oil, entirely free from animal or vegetable admixture, 
should be used for cylinder lubrication. With lubricant of this 
character modern boilers should be free from corrosion charge- 
able to the action of fatty acids. 

These are, however, not the only acids which have given 
trouble in the way of boiler corrosion. Under certain circum- 
stances free hydrochloric or muriatic acid is found in boilers. 
This is presumably due to the breaking up of magnesium 
chloride, forming hydrochloric acid and magnesium hydrate. 
The most dangerous feature of the corrosion due to hydro- 
chloric acid is that under conditions which may exist within 
steam boilers the chloride of iron first formed may become 


broken up, giving rise to other neutral compounds of iron, and 
setting free the acid to continue its ravages. 

There are also possibilities of the development of nitric acid 
from the organic matter which in small quantities may occa- 
sionally find its way into steam boilers. 

Except as it may be modified by electro-chemical action, 
the presence of such an acid usually results in a general surface 
corrosion, at least of all surfaces not protected by a sufficient 
layer of lime scale. 

The most troublesome feature of boiler corrosion has not 
been, however, a general or more or less uniformly distributed 
effect, such as would naturally be charged to the action of an 
acid diffused throughout the boiler. It has been rather in the 
so-called pitting. This term refers to the formation of small 
pits or depressions from the size of a pin head upward, and 
conical or cup-shaped in form. The depth of such pits may be 
anything from a slight depression to a hole cut entirely through 
a boiler tube. They are found in no fixed locality, though more 
commonly on the tubes, furnaces, and combustion chambers 
than elsewhere. When found they are usually filled with 
a blackish or brownish pasty mass, consisting chiefly of iron 
oxide with a slight admixture of lime salts, oily matter, and 
other substances. This deposit within the pits is often covered 
with a skin of somewhat different composition, consisting of 
lime salts and iron oxide in more nearly equal proportions. 

To account for the formation of these pits, various explana- 
tions have been suggested, most of them involving electro-chemi- 
cal action as a more or less pronounced feature. To under- 
stand the nature of this action a few explanations must first be 

Nearly all substances are in a different electrical condition, 
or at a different electrical potential, as it is called. This differ- 
ence is found not only between substances of different kinds, 
but also between similar substances at different temperatures, 
or in different physical conditions, as, for example, between two 
pieces of iron or steel, one of which has been hammered or 
worked more than the other. Due to this difference of elec- 
trical potential there is a tendency to set up a flow of electricity 
from one to the other, and as a further result to so change the 
two substances as to bring them into electrical equilibrium. In 
other words, the result of such a difference of electrical con- 


dition is always to bring about changes which will cause the dif- 
ference to disappear, and so bring the two substances into 
equilibrium. These chemical changes of the two substances, 
which tend toward electrical equilibrium, may be much helped 
or hindered by the medium in which the bodies are immersed. 
If they are in dry air, for example, no such activity takes place, 
and the difference of electrical condition continues unchanged. 
If, however, they are immersed in water, or especially in salt or 
slightly acid water, the operation will usually be much assisted 
by the activity of the medium for the substances. It may also 
happen that the medium and substances are so related as to 
bring about a series of chemical changes, of which the first are 
those which would naturally be associated with the transfer of 
electricity and the development of equilibrium, while the second 
counteract these changes chemically, and bring the substances 
back to their original condition, and so keep them constantly in 
the condition of electrical difference. There is as constantly 
the attempt to restore equilibrium, and hence so long as these 
conditions continue there will result this continued series of 
chemical actions, accompanied by a constant flow of electricity 
from one substance to the other. In order, however, that this 
flow of electricity may be thus constant and so constitute a cur- 
rent of electricity, as it is termed, there must be a path for a 
complete circuit or flow in one direction through the medium 
which produces the chemical changes, and in the other direction 
outside of this medium. The substance from which the current 
flows in the medium is known as the electro-positive element, 
and the other the electro-negative. The chemical activity pro- 
ceeds and the current is formed, in general, at the expense of 
the electro-positive element. 

These operations are illustrated in the ordinary voltaic cell 
or battery, such as those used for ringing bells, etc. In most 
of these batteries, however, the action is not self-sustaining, and 
if allowed to continue for a little time, a condition of electrical 
equilibrium is reached, or, as ordinarily stated, the battery is 
run down. In others used for telegraphy and other purposes 
the operations are self-sustaining and continuous until the 
chemical substances are exhausted. 

In a boiler these conditions for a more or less continued 
electro-chemical action may be fulfilled in a variety of ways. 
Parts of the structure of widely differing temperatures or of 


different physical or chemical compositions may provide the 
elements in a different electrical condition. Still more likely 
is such a difference to be found between iron and its oxides, 
especially the magnetic oxide or mill scale (Fe 3 O 4 ), or between 
a particle of carbon in the steel and the surrounding metal, or 
between a place in the steel where the proportion of carbon is 
much greater than the average and the surrounding metal, or 
between a bit of slag or other impurity in wrought iron and the 
surrounding metal. Copper, either in the form of oxide, or es- 
pecially in the metallic form, would also supply a substance dif- 
fering strongly from the iron. The exciting liquid is the water 
in the boiler, and its action will be more vigorous according as 
it is more acid in reaction, higher in temperature, and denser in 
concentration. With a high pressure boiler, water of high 
density and quite acid in character, and with the usual lack of 
homogenity or uniformity in the structure of the boiler, we 
should, therefore, expect the effects of electro-chemical action 
to be shown in marked degree. It happens, furthermore, that 
iron is electro-positive to copper, to carbon, and to its own 
oxides, so that in all cases likely to occur the operation will 
proceed at the expense of the iron. . 

From the very nature of these electro-chemical actions 
their effects are necessarily local in character, and so far as 
understood they seem to provide a fairly good explanation of 
the formation of pits as already described. It is not unlikely, 
however, that in some cases they are due rather to simple 
chemical action, and that their localization to a small spot is 
due to special or accidental causes, such as the protection of the 
surrounding'metal by lime scale, or a peculiar weakness against 
chemical attack at that point, due to peculiarities in chemical or 
physical structure. 

The possibility of deposits of copper on boiler surfaces has 
been already mentioned. These were first noted in connection 
with the corrosion accompanying the general introduction of 
the surface condenser. It was believed that the copper of the 
condenser tubes was attacked by the pure water resulting from 
the condensation of the steam, or by the fatty acids formed as 
above explained, and was then carried over into the boiler and 
deposited on the surfaces. To prevent such action the con- 
denser tubes were tinned, thus covering the copper from the 
action of the water or the fatty acids. Neither this step nor 


the substitution of hydrocarbon oil for that containing fatty 
acids has made any very marked difference in boiler pitting, and 
at the most the presence of the copper can have been only one 
among a number of causes as suggested. 

There has been much difference of opinion and difference 
in experience regarding the question whether wrought iron or 
steel boiler tubes were the more liable to corrosion. It was 
pointed out that wrought iron was less homogeneous than steel, 
and therefore the latter should be the better. The early experi- 
ence with steel hardly bore out this claim, and in fact the gen- 
eral opinion seems to have been that wrought iron tubes were 
found to corrode less readily than steel. In explanation of this, 
it may be said that while wrought iron was less homogeneous 
physically, the steel was perhaps less homogeneous chemically, 
and in any event contained a larger proportion of carbon than 
the iron, so that it would by no means follow that it would neces- 
sarily be less subject to electro-chemical action. The latest and 
best products of the steel makers for such purposes, however, 
are extraordinarily low in carbon and very homogeneous, and 
experience with such grades of material seems to show them 
superior to wrought iron in this respect. 

We have thus developed in some detail the causes of corro- 
sion on the water side of steam boilers, so far as they are under- 
stood. For the prevention of such effects their causes must be 
removed or counteracted. 

For reducing the amount of oxidation and the possible re- 
sults due to electro-chemical action, the presence of air in the 
feed water must be avoided by preventing as far as possible the 
entrance of water from overboard into the feed. The hot-well 
or feed-tank should also be of good size and kept full, so that 
there may be no danger of its getting too low 7 from time to time, 
and thus allowing the pump to take air. The piston rod on the 
low-pressure cylinder should be kept well packed, so as to pre- 
vent the entrance of air during the exhaust part of the stroke. 
The feed pump rods on the water end should be kept well packed 
for the same reason. 

To prevent acid corrosion the formation of the acids must 
be prevented as far as possible, and such as may form must be 
neutralized within the boiler. The prevention of the formation 
of fatty acids has been considered above. We have also seen 
that the formation of other acids is due chiefly to the presence 


of salts contained in sea water, or to organic substances. We 
have therefore simply an additional reason for keeping all such 
substances out of the boiler as far as possible. To neutralize such 
acid as may form, bicarbonate of soda, or soda-ash, as it is 
known in the trade, may be used from time to time, and in such 
quantities as may be found necessary. To test the water for 
acidity the litmus test is used. Blue litmus paper turns red when 
dipped in water slightly acid, while if the water is alkaline it re- 
mains blue, or the red color caused by an acid is changed to 
blue. By this means the condition of the water may be tested 
from time to time and soda used accordingly. Care must be 
taken not to use it in too great excess, as it may cause foaming. 
The soda is introduced by means of a soda cock on the con- 
denser. Instead of keeping the water alkaline by the use of 
soda, dependence is often placed on the zinc slabs used to pre- 
vent electro-chemical corrosion. These are gradually dissolved, 
forming zinc chloride, and this will undoubtedly tend to neutral- 
ize free acids and to keep the water alkaline. Whether sufficient 
or not, can of course be readily determined by the litmus test 
before referred to. 

For the prevention of electro-chemical action the causes 
must also be removed or neutralized as far as possible. This 
cannot be realized entirely, but it is clear that the results will be 
the better, as the following conditions are the more nearly ful- 
filled : 

(1) The structure of the boiler should be of material as 
homogeneous as possible in its chemical constitution and physi- 
cal condition. 

(2) Causes liable to produce oxidation or the presence of 
foreign substances should be kept out of the boiler as far as 

(3) The water in the boiler should be made as nearly neu- 
tral or non-exciting relative to the iron as possible. This in a 
general way will be attained by keeping it slightly alkaline rather 
than acid, and by avoiding very high densities. 

In addition to these means for reducing the causes, there 
remains one further step, and that is : 

(4) The provision of a substance which shall be electro- 
positive to iron, and readily attacked, so that the activity will be 
diverted from the iron to the protecting substance, and the 
operation will proceed at the expense of the latter rather than 


of the former. Such a substance we find in zinc, and its use for 
this purpose is very general and seemingly beneficial. 

It may be also noted that the formation of zinc chloride as 
referred to in the foregoing will aid in keeping the water alkaline 
in reaction, thus reducing its natural activity, and contributing 
further to the general decrease of electro-chemical action. 

In order to be effective as a protection to the iron in the man- 
ner described, the zinc must be in actual metallic contact with 
the structure of the boiler. It is usually in the form of rolled or 
cast slabs, weighing 8 to 12 Ibs. each. These are often placed in 
perforated sheet metal baskets hung from the stays or attached 
to other portions of the boiler. Where the basket is attached to 
the boiler there should be bright metal contact, and the attach- 
ment should be by screwed joint or other equivalent means, so 
that the separation of the two surfaces by the formation of scale 
or corrosion between them may be prevented. The zinc should 
also be connected to the basket by through bolts or other means 
which will insure continuous metallic contact. In some cases 
the zincs are hung by a through bolt without other means of sup- 
port. In such case, as the zinc becomes used it may fall apart and 
the pieces may lodge where they will obstruct the circulation, or 
be otherwise undesirable. In any event, they will no longer pro- 
tect the part of the boiler confided to their care, and their period 
of usefulness may therefore be less than when supported and 
connected to a basket, as described above. The number of zincs 
fitted varies greatly, according to the judgment of different engi- 
neers. In some cases not more than 10 or 12 would be assigned 
to the protection of a large double-end boiler, while in others as 
many as 40 or 50 would be used. The latter number is the bet- 
ter representative of good modern practice. In any case they 
should be distributed as nearly uniformly as possible throughout 
the boiler, in order that the latter may be thus subdivided into 
parts, each more especially, under the influence of a given slab. 

In connection with the use of zinc it may be noted that for 
such boilers as may be used for distilling purposes, that is, for 
the provision of fresh water for drinking and cooking, the zincs 
should be omitted, as the presence of any considerable amount 
of zinc chloride will render the water unsuitable for such uses. 

Instead of depending on zinc to prevent or divert electro- 
chemical action, as above described, some engineers prefer to 
depend simply on reducing the activity of the water by keeping 


it alkaline by the use of soda, introduced as the litmus test may 
show to be necessary. 

When spots are found in a boiler, showing the presence of 
pronounced corrosion, they should be cleaned off thoroughly, 
washed with soda solution, and, if not on a heating surface, cov- 
ered with a thin wash of Portland cement. This will attach itself 
to the iron and protect it in a manner similar to the lime scale. 

The beneficial effect of scale in thus protecting the surfaces 
of boilers from corrosion is well recognized, and there is no 
doubt that its presence as a thin wash or layer is of great value. 
In order to be effective, however, it must be so firmly and closely 
attached to the iron as to prevent contact of the water with the 
surface, else the corrosive action may proceed under the scale 
and result all the more seriously because it is protected from in- 
spection until the scale is thoroughly removed. On the tubes 
and other heating surfaces of the boiler, with their changes of 
temperature and consequent expansions and contractions, the 
scale is especially liable to be cracked off or partially separated 
from the iron, with possible results, as here noted. This is still 
more likely to be the case as the scale becomes thicker and the 
metal more liable to become overheated. It has also been sug- 
gested that a very heavy scale may result in an overheating of 
the metal sufficient to decompose the moisture present, thus lib- 
erating oxygen and forming the magnetic oxide of iron or black 
mill scale (Fe 3 O 4 ). This is highly electro-negative to iron, and 
thus it may give rise to harmful electro-chemical reactions. 

Laying Up Boilers. When boilers are to be laid up, the prin- 
ciples already explained will indicate the nature of the means 
suitable for preventing corrosion. 

On the outside, paint or other like coating may be used, as 
already noted. On the fire side of water-tube boilers protec- 
tion is sometimes gained by building a slow fire of tar or resin- 
ous material, the> tarry smoke from which condenses on the 
tubes and furnishes protection from the air with its moisture and 
carbon dioxide. Use is also made of quicklime in trays renewed 
from time to time. This absorbs the moisture and so keeps the 
air dry. 

On the inside, all boilers when laid up should be either 
empty or entirely full. If a boiler stands for any considerable 
length of time partly full, corrosion is likely to occur about the 
water line. If they are to be out of use for a short time only, 


they may be filled full of water made slightly alkaline by the ad- 
dition of soda, the condition of the water being determined by 
the litmus test already referred to. If they are to be laid up for 
a longer time it is better to lay them up dry. To this end the 
water is removed, the manhole-plates taken off and the interior 
thoroughly dried out by the introduction of trays of burning 
charcoal or coke. The boiler is then closed up, except a lower 
manhole, through which a tray of freshly burning charcoal is in- 
troduced, and the manhole cover is put on. The charcoal will 
consume most of the remaining oxygen, and the boiler will thus 
be protected. Instead of the final introduction of a tray of char- 
coal, trays of quicklime may be used to insure the absence of all 
moisture, and the boiler then closed as before. 

It is readily seen that these various methods are simply 
ways of carrying out the necessary conditions for preventing 
oxidation, as already discussed, and if these principles are kept 
clearly in view the means most conveniently at hand may be 
suitably adapted to provide the protection desired. 

Sec. 41. BOILER SCAI,E. 

It is well known that sea water contains in solution a certain 
amount of solid matter, while even ordinary fresh water is not 
wholly free from similar substances. As long as the water re- 
mains in its natural condition these solids remain in solution ; 
but under the change of condition to which the water in a steam 
boiler is subjected, they are liable, as explained later in detail, to 
separate out from the water and thus to form scale or sludge, 
according as the circumstances may determine. 

The proportion of the solid matter in ordinary sea water is 
about (by weight) i part in 32, or 1-32. This is the same as 
about 5 oz. per gallon, or 2 Ibs. per cu. ft. The solid matter con- 
sists chiefly of chloride of sodium or common salt, with small 
quantities of calcium sulphate and carbonate, magnesium sul- 
phate and chloride, with smaller quantities of other substances. 
An average composition of this solid matter is about as follows : 

Chloride of Sodium (common salt) 76 per cent. 

Chloride of Magnesium 10 

Sulphate of Magnesium 6 

Sulphate of Calcium (gypsum) 5 " 

The remaining 3 per cent consists of small quantities of 
other salts with a little organic matter. 

The proportion of the solid matter in river and lake water is 


quite variable with the locality, and no representative or average 
analysis can be given. The amount held in solution may vary 
from perhaps 10 to 250 parts in 100,000 or from .015 oz. to .30 
oz. per gallon, or .1 oz. to 2.5 oz. per cu. ft. It is composed 
chiefly of the carbonates of calcium and magnesium with smaller 
quantities of the sulphates of calcium and magnesium, and other 
substances. In addition to the substances in solution, quanti- 
ties of sand, mud, organic matter, etc., may be carried in sus- 
pension, dependent entirely on the locality and special circum- 

Boiler scale from sea water is composed chiefly of calcium 
sulphate or sulphate of lime, as it is commonly called, while that 
from fresh river or lake water is composed chiefly of calcium 
carbonate or carbonate of lime, as commonly called. With 
brackish water, as we might expect, the proportions of the two 
are more nearly the same. Following are analyses of boiler scale 
by Professor Lewes which may be considered as typical of the 
incrustations formed by river water, brackish water and sea 
water, respectively : 


Calcium Carbonate 75-85 

Calcium Sulphate 3-68 

Magnesium Hydrate 2. 56 

Sodium Chloride 0.45 

Silica 7.66 

Oxides of Iron and Alumina 2.96 

Organic Matter 3.64 

Moisture 3.20 

100.00 100.00 100.00 

It thus appears that scale from river water may be looked 
on as an impure calcium carbonate, that from sea water as an 
impure calcium sulphate, while that from brackish water, as we 
should expect, is a mixture of the two in more nearly equal pro- 

Sodium chloride or common salt is soluble in water until 
the proportion exceeds some 25 or 30 per cent. This corre- 
sponds to a density of 8 or 10 on the usual hydrometer, and is far 
greater than that reached by the water in marine boilers. This 
substance therefore gives no trouble so far as helping to form 
scale is concerned, and the small amount found in analysis of 
boiler scale is probably due to the shutting in, so to speak, of a 
small amount of water during the formation of the scale. In 
discussing the formation of boiler scale for our present purposes, 


it will be sufficient to refer to the behavior of the salts of calcium 
and magnesium. 

Calcium carbonate (CaCO 3 ) is practically insoluble in water, 
while calcium bicarbonate (CaC 2 O 5 ) is quite soluble, and it is in 
this form that the substance exists in solution in water. If r?ow 
the \\ater is heated to the boiling point carbonic acid (CO 2 ) is 
driven away from the bicarbonate, it becomes reduced to the 
simple carbonate, and being now insoluble it separates out a? a 
more or less powdery deposit. Mixed with other salts, how- 
ever, especially calcium sulphate, or if there is a little sulphuric 
acid in the water, it may collect on the heating surfaces and 
form a hard and closely adhering scale. Magnesium bicarbonate 
is in a similar manner reduced to the simple carbonate, which is 
insoluble, and is then deposited in the same fashion. 

Calcium sulphate is soluble in cold water to a slight extent, 
as found in sea water. As the water is heated, however, or as 
the density becomes greater, the proportion of sulphate which it 
can retain in solution becomes less and less. When the tempera- 
ture rises to 280 or 290 (corresponding to from 35 to 45 Ibs. 
gauge pressure) the water can no longer retain any of the sul- 
phate in solution, and it is all deposited. It is also largely de- 
posited, even at a temperature of 212, if the density rises to 
3-32 or above. The other sulphates become likewise insoluble 
and are completely deposited if the temperature rises to about 
350 or over, corresponding to about 120 Ibs. gauge pressure.' 
These sulphates of lime and magnesium thus deposited tend to 
attach themselves to the surfaces within the boiler, and to form 
a very hard and crystalline scale. 

As to the effects of this scale, its presence in a very thin 
layer is often considered beneficial as a protection to the surface 
of the boiler from corrosive influences. On the other hand, how- 
ever, it is a much poorer conductor of heat than metal, and its 
presence on the heating surfaces retards the transmission of heat 
from the fire through to the water. In the extreme case the heat 
may be so effectually shut off from the water that it simply be- 
comes banked up, so to speak, in the metal, and in this way the 
tubes and other heating surfaces may become seriously over- 
heated with resulting damage to the boiler. The scale may also 
in extreme cases become so collected between the tubes or be- 
tween the combustion chamber and boiler sheets as to impede 
the circulation of the water and thus lead to overheating and its 


dangers, as referred to above. In water-tube boilers the accu- 
mulation of scale on the inside of the heating tubes is of special 
danger, as the circulation becomes in such case rapidly ob- 
structed and the danger of overheating and rupture is corre- 
spondingly increased. In a similar manner the accumulation of 
scale in the interior of tubular feed-water heaters rapidly de- 
creases their efficiency as heaters, if no worse results follow 
due to the burning out of coils, or to the resulting shortness of 
water in the boilers. 

Scale Prevention, Fresh Water. The only sure way of pre- 
venting scale is simply to keep it out of the boiler. If the scale- 
forming substances find entrance to the boiler it will be found 
very difficult to prevent its formation, at least to some extent. 

On boats navigating inland waters the jet condenser is still 
for the most part used, the feed is ordinarily taken from 
the condenser, and therefore practically from overboard. 
In such boilers, therefore, we may expect the formation of the 
usual fresh water scale, consisting chiefly of calcium carbonate. 
For the treatment of fresh water scale a great variety of methods 
have been proposed. In some cases the substances proposed act 
chemically, in others mechanically. From the great variation in 
the character of the solid matter contained in fresh water, it can 
hardly be expected that any one method of treatment or sub- 
stance will prove equally good in all cases. 

If a feed-water heater is used, and is effective in heating 
the water, it will be found that most of the scale will be deposited 
in the heater, especially if it is of sufficient size to allow of proper 
time. In this way the scale may be kept out of the boiler proper. 
The heater, .however, should be so made as to readily admit of 
cleaning, especially if the water contains any considerable pro- 
portion of scale-forming salts, otherwise it will soon become 
choked and ineffective. 

Among the various substances which have been recom- 
mended for the prevention of fresh water scale the following 
may be mentioned : 

Oak and hemlock bark and other like substances which 
contain tannic acid are more or less effective in waters contain- 
ing carbonates of calcium or magnesium. The tannic acid, how- 
ever, will attack the iron of the boiler and may lead to serious 

Molasses, cane juice, fruits, distillery slops, vinegar and 


other like substances containing acetic acid have also been used 
with success where no sulphates are present. The acetic acid, 
however, is still more injurious to the iron than the tannic acid, 
and the organic substances will form a scale with sulphates if 
they are present. 

Sal-ammoniac when used with a feed water containing cal- 
cium carbonate brings about an exchange between the two sub- 
stances as a result of which ammonium carbonate and calcium 
chloride are formed. The former of these is soluble and quite 
volatile and passes off mostly with the steam. The latter is quite 
soluble and thus the deposition of the calcium carbonate is 
avoided. This operation by itself, however, would result in the 
gradual accumulation of calcium chloride in the boiler, thus rais- 
ing the density of the water to a point where ultimately it would 
begin to deposit. This condition may, of course, be controlled 
by a suitable use of the blow. 

Tanate of soda is well recommended for general use, but 
with water containing sulphates a small amount of soda-ash 
should be added. 

Among the substances which act mechanically, crude petro- 
leum and kerosene oils are probably the most widely used. The 
latter may be recommended as the better of the two, as the crude 
oil will sometimes aid in scale formation. They seem to act best 
in cases where there are some sulphates present, as in slightly 
brackish water, or in the waters of certain geographical regions. 
Kerosene seems to act by preventing the particles of scale from 
sticking closely together or from tightly adhering to the heating 
surfaces, so that much of the matter will collect as a sludge in 
the bottom of the boiler, and that on the heating surfaces will be 
more easily removed. 

In all cases where there is reason to expect the accumula- 
tion in the bottom of the boiler of deposits thrown down in a 
loose or powdery form, the bottom blow should be freely used 
so as to prevent the accumulation of too great a quantity, or op- 
portunity for its hardening into scale. 

In spite of all modes of treatment there will be found some 
scale on the heating surfaces, and provision must be made for 
entering the boiler and removing it with appropriate tools as the 
occasion demands and circumstances permit. 

In many of our inland waters the amount of scale-forming 
substances is so small that no special treatment is thought neces- 


sary, and little attention is paid to the matter except to remove 
the accumulation at the periods of regular inspection and over- 

Scale Prevention, Salt Water. Turning now to boilers in 
which sea water may form a portion of the feed, it will be of in- 
terest to first note briefly the historical development of the mod- 
ern situation. 

In the early days of marine engineering, the temperature 
and pressure of the steam were low, and the jet condenser was 
in general use. The feed water which was drawn from the min- 
gled condensing water and condensed steam was but slightly 
fresher than sea water, so that large amounts of solid matter 
were thus fed into the boiler. In consequence the density would 
have risen rapidly had it not been kept down by blowing off a 
part of the water in the boiler of relatively high density and re- 
placing it with the salt feed of lower density. Had the sulphates 
of calcium and magnesium thus brought into the boiler been 
completely deposited, enormous quantities of scale would have 
been formed, and this method of operation would have been 
quite impracticable. Due, however, to the moderate pressure 
then in use and to the fact that the density was kept usually be- 
tween i 3-4 and 2, the salts were held fairly well in solution, and 
but a moderate amount of scale was deposited. 

As steam pressures advanced, however, beyond 40 or 45 
Ibs., conditions were reached under which first the calcium sul- 
phate and later magnesium sulphate and other salts are com- 
pletely deposited. Under such circumstances blowing off to re- 
duce the density of the water will only make matters so much 
the worse,, for the lower the density is to be maintained the 
greater must be the amount blown off, and hence the greater the 
amount of extra feed, and the greater the amount of scale form- 
ing salts brought into the boiler, all of which will be deposited. 

It became therefore necessary to abandon the use of the jet 
condenser and salt feed. Its place was taken by the modern sur- 
face condenser. So long as this condenser is perfectly tight the 
feed water consists of the condensed steam, and is therefore al- 
most perfectly fresh water. Due, however, to steam leaks at the 
various joints, seams and glands, to the occasional use of the 
steam whistle, and to the use of steam in certain auxiliaries from 
which it is not returned to the condenser, there will be a con- 
tinual shortage in the feed water which under usual conditions 


will be found between say 2 and 5 per cent. Until recent years 
this shortage was made up by the use of sea water obtained 
usually by opening, as circumstances required, the salt water 
cock connecting the salt water side of the condenser with tne 
steam side. It is very difficult to keep the tubes oi a surface 
condenser packed perfectly tight, and in some cases the con- 
denser was allowed to run a little leaky, simply to make up in 
this way the salt feed required. 

Due to this admixture of salt feed, the scale forming salts 
of which are all deposited in the boiler, there will be a gradual 
formation of scale greater or less, according to the length of the 
run and the proportion of salt feed make up. 

In recent years experience has clearly shown that the dan- 
gers of overheating and the general bad effects due to the pres- 
ence of scale are more and more pronounced as the pressures 
are higher. It has become therefore more and more important 
to prevent so far as possible the entrance of any sea water into 
the boiler, and thus avoid the formation of scale with its troubles 
and dangers. To this end, in modern practice, the make up feed 
is provided by an evaporator, or in some cases by feeding one 
boiler with salt feed and thus restricting the scale formation to 
this boiler, while the condensed steam from all the boilers is re- 
turned to the other ones as feed. In all such cases it will be 
noted that this scheme amounts to a transfer of the use of salt 
water and the formation of scale from the boilers in general to 
the evaporator, or to the particular boiler in which it is allowed 
to accumulate. 

For short trips as, for example, those met with in bay, har- 
bor or channel service, or on short coasting voyages, fresh water 
for make up feed may be carried in tanks instead of providing it 
by means of an evaporator. By many engineers this is consid- 
ered the preferable method whenever tanks of sufficient size 
can be provided, and in some cases with the double bottom style 
of construction, double bottoms have been utilized to a consid- 
erable extent for this purpose. 

It is rare that the condenser can be maintained perfectly 
tight, so that even under the best practicable conditions there is 
apt to be some passage of sea water into the steam side of the 
condenser, and thence into the boiler. Under the best condi- 
tions the amount of scale formed, however, is so small that com- 
monly no special treatment is attempted, and the scale is allowed 


to deposit, and is then removed at the regular periods of inspec- 
tion and overhaul. 

Some attempts have been made to prepare sea water by the 
removal of the calcium sulphate in a separate vessel before en- 
tering the boiler. This may be done by the use of sodic fluoride 
which causes the sulphate to separate out and settle to the bot- 
tom as a fine powder. The remaining water is practically free 
from this substance and may be used for boiler feed without fear 
of causing scale. 

Soda ash and other alkalies have sometimes been used in 
boilers, with feed water containing sulphate of lime. They act 
by converting the sulphate into a carbonate, and thus into a 
somewhat less objectionable form. 

Barium chloride acts in a somewhat similar fashion by pro- 
ducing barium sulphate and calcium chloride. 

The use of zinc in boilers is also by many believed to prevent 
to some extent the formation of scale by the reaction of the alka- 
line zinc chloride on the scale forming salts. 

With sea-going, as with inland boilers, the bottom blow 
should be used occasionally and as the particular circumstances 
may demand, so as to remove the accumulation of such sub- 
stances as may be thrown down as a powder or sludge and thus 
collect in the bottom of the boiler. 

However careful the provisions for keeping sea water out 
of the boilers or no matter what methods may be used to pre- 
vent scale formation, it is almost sure to gradually accumulate, 
and assurance of safety from the troubles and dangers which 
may result can only be obtained from periodical examination 
and scaling as may be found necessary. All marine boilers must, 
of course, be provided with manhole plates for this purpose, and 
the internal arrangement of tubes, braces, furnaces, etc., should 
be made, so far as possible, with a view to furthering this neces- 
sary operation. 

Combinations of Oil and Scale. We have thus far referred 
to scale formed simply from the solid matter in the feed water. 
The combinations which may be formed by the deposited salts 
and oil from the cylinders as it may enter with the feed water 
are, however, of even still greater importance, and must now be 

Oil coming in thus with the feed water is caught by the cir- 
culating currents and distributed more or less throughout the 


bailer, though by reason of its lesser weight it will tend grad- 
ually to rise and accumulate as a scum at the surface of the 
water. In thus wandering about, a drop may come in contact 
with a bit of solid matter separated from the water. The two 
join together, the oil forming a coating about the sulphate, and 
they journey on meeting and joining with other like particles. The 
combination of the oil and sulphate may have about the same 
specific gravity as the water in the boiler, and hence these parti- 
cles will readily move with the circulating currents, either up or 
down, as they happen to be flowing. They are thus swept along 
the heating surfaces, to which they attach themselves all the 
more readily by reason of their oily -covering, and on either the 
upper or lower side as they happen to be moving with a down 
or up-flowing current. In this way the coating gradually in- 
creases until it has attained a thickness sufficient to seriously 
interfere with the passage of the heat. 

In other cases, when the scale and oil are lighter or the 
water is denser and heavier, there seems to be formed at the 
surface of the water in the boiler a kind of oil and scale blanket 
or layer floating about, and perhaps ultimately by the gradual 
increase of weight sinking and covering some portion of the 
heating surface. Especially is this oil "pancake," as it has been 
called, liable to settle should the density of the water in any way 
be suddenly decreased. Still otherwise should the boiler be 
blown down by the bottom blow, such an oil blanket would 
naturally settle and attach itself to some part of the heating sur- 
face. Should the boiler be then filled again, the coating would 
remain where attached. This shows that under such circum- 
stances a boiler should never be blown down with the bottom 
blow without first using thoroughly the surface blow to remove 
as far as possible all such accumulations of oil or of oil and 
scale from the surface of the water. 

The danger to be feared from this combination of scale and 
oil is not in its close adherence to the surfaces, but in its non- 
conductivity for heat. Experiments show that 1-16 to 1-8 inch 
of such a covering is far worse in this respect than perhaps 1-2 
inch or more of scale alone. The danger to be feared is there- 
fore overheating and collapse, and not a few cases of the col- 
lapse of furnaces and other parts of marine boilers are believed 
to be due to this cause. 

So far as these effects are concerned it is seen that it is bet- 


ter to carry a high density in the boilers than a low one, so as 
to keep such oil and scale combinations at the surface of the 
water, where they may be disposed of by the surface blow. 

As it is practically impossible to prevent the entrance of 
some scale-forming materials into the boiler, the danger of 
trouble with oil and scale combinations is most surely pre- 
vented by keeping the oil out. To this end a cylinder oil should 
be used having a high point of vaporization, as the higher this 
point the smaller the amount carried into the condenser. Of 
this oil the minimum amount necessary should be used in the 
cylinders, and the feed water should be filtered to remove what- 
ever oil it may contain. 


[i] Inspection and Test. 

We will suppose that after some considerable term of serv- 
ice, and preparatory to a general overhauling, a battery of 
boilers are to be carefully and thoroughly examined. The 
more important points may now be considered. 

(1) Furnace Fronts. The furnace fronts and doors may be 
found warped and cracked, and if this is the case to such an ex- 
tent as to interfere with the proper closure of the furnaces, or 
with the proper and convenient care of the fires, the necessary re- 
pairs or renewals should be made. 

(2) Grates and Bearers. The grate bars will often be found 
warped and twisted, or badly burned, and at various points sunk- 
en below or sprung above their proper level. Such irregularities 
in the grate may occasion loss of coal at some points, while they 
will further the accumulation of ash and clinker at others, and 
will make it almost impossible to give to a fire the proper atten- 
tion, or to get from a square foot of grate surface the power 
which it should be able to give. The bearers may also be the 
cause of trouble by warping or settling from having been over- 
heated, and all of these points must be attended to before the 
boiler can be considered again ready for proper service. 

(3) Bridge-wall The bridge-wall is liable to be found more 
or less burnt out and dilapidated, while on the front side clinker 
and bits of brick may be found fused together in irregular 
masses. All of this must be removed and the bridges built up 
again with fresh bricks to the proper height as referred to in 
Sec. 38 [i]. 


(4) Tubes. The tubes will, of course, be swept and properly 
cleaned on the fire side. The existence of small leaks must be 
carefully looked for, the evidence being the presence of soot and 
scale burned to the metal where the water has come through 
and evaporated. The back and front tube sheets and the inside 
of the tubes must be carefully examined for any such evidence. 
Signs of especial wear should also be looked for at the back 
ends of the tubes, and if ferrules are used some will probably be 
found so worn and burned out as to require renewing. 

In water-tube boilers any special warping or change in the 
shape or curvature of the tubes should be carefully noted, as it 
may indicate overheating due to faulty circulation caused by a 
clogging of the tube by scale and sediment. A split or badly 
ruptured tube will, of course, show itself by the resulting leak, 
but in a water-tube boiler such leak may be very difficult to 
locate without the removal of several of the tubes in the vicinity 
of the one giving the trouble. These points depend entirely on 
the type and style of construction, and no general rule can be 
given for definitely and immediately locating such a tube in a 
water-tube boiler. 

(5) Joints and Seams. The joints and seams throughout the 
boiler, both in the combustion chamber and on the outside, 
should be carefully examined for small leaks, either between the 
plates or about the rivets. If the leakage is not serious, caulking 
will serve as a sufficient remedy. In other cases, however, the 
removal of old rivets and the insertion of new ones may be 
found necessary. 

(6) Front Connections and Uptakes. The front connections, 
uptakes and fittings should be examined to make sure that the 
plates are not warped or broken from their fastenings, and that 
the dampers and their operating gear are in proper condition. 

(7) Fittings. The valves and cocks are likely to be found 
more or less worn on their seats and leaky in consequence. 
These will require regrinding and refitting, or replacing by new 
where necessary. The operating gear, such as valve spindles, 
wheels, levers, chains, gear-wheels, etc., should be examined for 
any breakage or derangement of parts. The various joints and 
fittings about the steam and water pipes must also be examined 
for signs of leaks, distress, corrosion, or other derangement. 

(8) Bracing. The manhole plates will, of course, have been 
removed to facilitate examination of the interior. 


The braces, especially where pin joints and like connections 
are used, should be carefully examined for defects in the con- 
nections and fittings, and also for any symptoms of buckling or 
distress in the braces themselves. 

(9) Scale. The scale present in the boiler should be ex- 
amined as to its amount, distribution and character whether 
hard or soft, greasy or otherwise, closely adhering or readily 
cracked off. Accumulation of scale between the tubes or screw 
stays, and of scale and sludge in the bottom of the boiler must 
also be looked for and noted. In some cases an oily or greasy 
coating with little or no mineral matter and forming a coating 
over the scale and on the heating surfaces may be observed. 
This will indicate large quantities of oil in the boiler, and insuffi- 
cient use of the surface blow. 

(10) Corrosion. It is, of course, of the highest importance 
to examine carefully for signs of corrosion and pitting through- 
out the interior of the boiler. The following locations, however, 
are those in which it is most apt to be found : 

On the sheets at and near the water line. Occasionally also 
severe corrosion is found in the steam spaces. 

On the braces near the water line. 

On the tubes and combustion chamber tops. 

On the furnaces near the grate level. 

The nature and distribution of this corrosion must be care- 
fully noted, in order that the most suitable steps may be 
taken for its arrest and prevention in the future. When they 
can be gotten at, corroded spots may be scraped and scrubbed 
clean with water made alkaline by the addition of soda or weak 
lye, and if not on a heating surface, a redistribution of the zincs 
may prove of service, while in general a more careful attention 
to the various means suggested in Sec. 40 may be recommended. 
The zincs and their fittings., as discussed in Sec. 40, must also be 
carefully looked after. Many of the zincs will probably be found 
to have wasted away to only a small part of their original size, 
and to have become changed in physical structure to a blackish 
or brownish crumbly or brittle mass. In some cases remnants 
of the slabs may be found lodged between the tubes and screw 
stays, and often more or less covered or imbedded in deposits 
of scale. 

On the exterior of the boiler the points most liable to 
corrosion are on the fronts about the bottom where damp ashes 


may have lain, or about the saddles and on the under side where 
dampness and water are liable to be formed. Thorough clean- 
ing, followed by a coat of paint, asphaltum varnish, or other 
like material, is the usual remedy in such cases, at least where 
its application is practicable. 

Before the application of any such coating, the plates should 
be thoroughly dried, else it will be of little use. The presence 
of moisture on the plates causes the especial difficulty connected 
with the effective application of paint in such places, and where 
convenient the use of a portable sheet-iron drying stove contain- 
ing burning charcoal or coke may be found of use. This may 
be placed under the surfaces to be covered so as to furnish an 
ascending current of warm air, thus aiding in keeping them dry 
during the application of the paint. 

For the structural material in bilges and bunkers a coating 
of Stockholm tar put on hot and then sprinkled with Portland 
cement is highly recommended by some engineers. 

(n) Manholes and Covers. The faces on which the manhole 
cover joints are made should be examined for corrosion or 
scale, or anything which may affect their evenness, or make 
difficult the fitting of a tight joint. 

(12) Drill Test. Where the boiler has seen long service, or 
where there are evidences of serious corrosion, or doubt exists 
as to the thickness or quality of the plates, they must be drilled 
at such points as may be selected. In this way the thickness of 
the remaining good metal may be ascertained, and the safe 
pressure to be carried may be fixed in accordance with the evi- 
dence thus found. 

(13) Hydraulic Test. When the boiler has been overhauled 
and put in proper condition, at least as far as anything which 
may affect its strength is concerned, the hydraulic test may be 
applied. To this end the boiler is filled full of water and pres- 
sure is put on, usually by means of a special pump connected 
for the purpose. The test pressure is usually one and one-half 
times the working pressure desired. It is considered that this 
pressure is not sufficient to seriously try or injure the boiler 
should it be properly constructed, and of suitable factor of safety 
throughout, while at the same time it will be sufficient to de- 
velop small leaks, and should the boiler be unduly weak at any 
point, the bulging or yielding or distress at such point should 
become apparent. If no such evidences appear, then it is con- 


sidered that the boiler is abundantly strong for the working 
pressure as desired. 

To prepare the boiler for the test the springs should be 
withdrawn from the safety valves and lengths of pipe of suitable 
size substituted, so that the valves may be screwed down fast. 
All stop valves and gauge glass connections should then be 
tightly closed, as well as the connection to any pressure gauge 
which will not indicate up to the test pressure. 

While the test is under way the boiler is subjected to the 
most careful examination, both inside and out. The furnaces, 
combustion chambers and back tube-sheets are examined from 
the inside, and the shell and its various joints and seams from 
the outside. Small leaks are watched for and stopped by caulk- 
ing, if possible, or if about the tube ends, by re-expanding. Es- 
pecial care must also be had in watching for any signs of bulging, 
buckling or other deformation or distress. 

Where the test is to be carried out with especial care, ex- 
tension and compression gauges are provided in the furnaces 
and combustion chambers, and at other points, as may be de- 
sired. These serve to indicate and to measure the actual amount 
of distortion which results from the gradually increasing pres- 
sure. If the distortion or bulging at any point should become 
abnormal, the pressure should be relieved by letting off a little 
water, in order to see if any permanent set has been made. 
The continuance of the test should then be made to depend 
upon the behavior of the part showing this relative weakness. 
For a thoroughly satisfactory test, all gauges should return to 
the original setting when the pressure is removed, showing no 
permanent set at these points.* 

It will usually be found very difficult to so tighten the vari- 
ous valves that there will be no leakage. An idea of the amount 
of leakage may be obtained by watching the rapidity with which 
the pointer on the pressure gauge moves backward when the 
pump is stopped, as well as by the amount of pumping required 
to maintain the pressure at its full value. The pressure gauge 
pointer will also frequently indicate by its more or less sudden 
movement backward, the sudden development of leaks about 
the riveted joints or tube ends. 

It has sometimes been urged against the hydraulic test that 
it may severely strain some part of the boiler where the yield 
or distress is difficult to observe, and thus so weaken it that a 


further yield or rupture may occur under a much smaller, load 
at a later time. The hydraulic test is, however, very generally 
employed both in the naval and mercantile marines, and if care- 
fully conducted and with a pressure not exceeding one and one- 
half times the working pressure, it is not likely that harm will 
result, while the test will develop the small leaks and minor 
defects, and may be the means of exposing serious faults of 
workmanship or design. 

The same test is, of course, applied to new boilers as a 
final preliminary to the getting up of steam. 

In some cases the water test has been carried out by filling 
the boiler and then lighting wood fires within the furnaces. The 
expansion of the water will furnish the increase of pressure de- 
sired, which may be eased by the safety or stop valve, as neces- 
sary. It has been claimed that the boiler being in this way 
heated, was more nearly in its regular service condition. While 
this may be so to a slight extent, the boiler is, nevertheless, far 
from regular service condition, and the method has the serious 
disadvantage that it does not allow examination of the furnaces, 
combustion chambers and back tube sheets while it is under 
way. It is also under less ready control than the pump method, 
and is now but rarely employed. 


In the following suggestions regarding boiler repairs we 
shall refer more especially to such as may become necessary at 
sea or under emergency conditions, rather than to those which 

may result in the course of a thorough overhauling in port. 

(2) leakage from the Joints of Boiler Mountings. 

Such leakage by soaking through the lagging and keeping 
the plates wet may give rise to surface corrosion on the boiler 

The first care must be to stop the leakage by screwing 
down, re-caulking or re-making the joints, as may be necessary. 

If there is reason to suspect corrosion of the boiler as well, 
the lagging should be removed and the corroded surfaces 
scraped clean and painted with good metal paint or other suit- 
able covering. 

(3) Leakage About Shell Joints. 

Usually caulking will be sufficient to stop any ordinary 
small leak in these joints. If it is serious and caulking gives 


but Jittle improvement, it may indicate a loose rivet or one with 
the head gone. In such case leakage about the rivet will usually 
be present also, and will thus serve to locate the trouble. Such 
rivet must, of course, be replaced, in order to effectually stop 
the leak. 

Where a rivet has blown out and a quick repair is desired, 
the hole may be drilled or reamed true and then tapped out. 
Then fit a bolt with corresponding thread and cut it partly 
through near the root with a hack saw. Screw this in, knock 
off the projecting end, rivet down the remainder and the job 
is complete. 

In some cases instead of replacing loose or broken rivets, 
or where for other reasons caulking seems to be inefficient in 
stopping the leak, it may be considered desirable to put a patch 
over the seam, rivets and all. In such case a so-called "soft- 
patch" is applied. This is illustrated in Fig. 204. The patch 
is flanged and made with a recess of suitable size and form to 
accommodate the rivet points. It is then filled with a stiff putty 


Fig. 204. Patch for Leaky Joint. 

of red lead and secured by bolts as shown. Such an application 
is really a red-lead poultice, kept in place by a suitably formed 
steel cover, and secured to the shell, as explained. It is un- 
necessary to make such a patch of metal more than 3-16 or J4 
inch thick, since it is not intended to add strength to the shell, 
but simply to keep the red-lead putty in place, and thus stop 
the jet of leaking steam or water. 

The chief difficulty with leaks in the shell seams and with 
the outside of boilers in general arises from the trouble in sub- 
jecting them to the proper examination due to the presence of 
the lagging. As usually fitted, this covering is difficult of re- 
moval, and small leaks thus covered in may continue for long 
periods of time, keeping the outer surfaces wet and causing 
rust and corrosion where its existence may not be expected. 
A form of boiler lagging admitting of ready renewal and re- 
placement in sections is much to be desired, and if full advantage 
were taken of such a form of covering to keep closer watch of all 


joints on the outer surface, much trouble might be avoided by 
taking the first appearances of trouble in time. 

(41 I/eakage at Internal Joints. 

The internal joints, on the whole, give more trouble than 
those on the outside. This is only to be expected due to the 
thinner plates, the enormous range of temperature differences 
which exist, and the resultant expansions and contractions. The 
difference in expansion between the furnaces and tubes is es- 
pecially liable to give trouble with the joints connecting the 
furnace to the combustion chamber, and several varieties of 
joint have been proposed to reduce this trouble to a minimum. 
With a joint such as shown in Fig. 43, the greater expansion of 
the furnace tends directly to open up the joint, while with that 
shown in Fig. 9 the result is a shear on the rivets but no direct 
tendency to open the plates. In the latter case, however, the 

Soft Patch. 

Locomotive Patch. 

Hard Patch. 
Fig. 205. Different Forms of Patches. 

rivets are more directly exposed to the fire than in the former, 
and thus the points between the two joints are very nearly 
balanced. The joint of Fig. 9, however, allows a more ready re- 
moval of the furnace, and on this account it is often selected 
rather than that of Fig. 43. 

Leaky joints on the combustion chamber should be first 
carefully re-caulked. This operation, however, cannot be car- 
ried on indefinitely, for after caulking to a certain extent, the 
edge must be chipped off to get a fresh caulking edge. This, 
if repeated, will leave the metal between the rivets and edge too 
narrow for safety. If careful and judicious caulking does not 
remedy leaks in these seams, it is evident that the rivets need 
renewal, and in carrying this out especial care should be taken 
to see that the holes are fair in the two plates, and that the 
rivets fill them completely. 


(5) Patches. 

We may now turn more especially to the patching of 
boilers and to the different kinds of patches employed. 

We must first remember as a general principle that any 
thickening of metal on the heating surfaces is undesirable and 
to be avoided as far as possible, or reduced to the lowest possible 
extent. If, therefore, a patch on a heating surface is to be 
considered as a permanent fixture, the faulty metal should be 
cut out, thus doubling the thickness only over the necessary 
width for the fastenings. A patch put on in this way with rivets 
headed up as in regular boiler work is known as a hard patch, 
and is illustrated in Fig. 205. In some cases the patch must 
be put on from one side only, or is more temporary in character. 
In such case either the locomotive or the soft patch is used. 
The former is a patch put on with tap bolts, as illustrated in Fig. 
205, and usually without cutting out the metal. The soft patch 
has been already referred to. In its more usual form, as shown 
in Fig. 205, it is made by lipping down a plate of steel so as to 
contain and hold in place a coating or layer of red-lead putty. 
It is more suitable for temporary repairs, or where the sur- 
faces are so rough and uneven that a patch of the other forms 
could not be fitted. The soft patch is sometimes secured with 
tap bolts, and sometimes with through bolts and nuts, as may be 
most convenient with the case in hand. 

As to whether a patch should be put on the water or fire 
side, much will depend on location and convenience. Where 
it is possible the water side may be chosen so that the steam 
pressure will tend to keep the patch in place. Usually, however, 
the fire side of the plate is more easily gotten at, and in many 
cases there is no choice but to put the patch upon this side. In 
any event with either the hard or locomotive patches, the edge 
will require careful caulking as the final closure and making of 
the joint. With the soft patch, caulking the edge is not neces- 
sary, as the putty is depended upon to stop the leak, and the 
office of the patch is merely to hold it in place. 

(6) Cracks and Holes. 

A small crack is usually treated by drilling a hole at each 
end to prevent its extension, and then covering with a patch, 
according to location and convenience. Very small cracks are 
sometimes drilled and tapped out as close together as the holes 


will stand, and then filled with soft iron or steel bolts, riveted 
down so as to overlap and thus completely close the crack. 

Small isolated holes not accompanied by a general thinning 
of the metal may be treated in a similar fashion by drilling and 
tapping out the hole and riveting in a bit of soft iron or steel 
bolt. Larger holes, or where many are located near each other, 
or where they are accompanied by pronounced thinning of the 
metal must be treated by patching. 

In general it may be noted that plugging as above de- 
scribed, is only suitable for the mere stopping of a leak, and 
that it adds nothing whatever to the strength of the plate. If, 
then, the conditions are such as to make additional strength de- 
sirable, a patch must be fitted. 

Where a crack is found in a tube sheet it usually extends 
from tube to tube. Such a crack may be covered by a patch ex- 

Fig. 206. Patch for Boiler Tube Sheet. 

tending over the crack and taking enough good metal to obtain 
a secure hold. The patch must have holes cut in it, of course, 
to correspond to the tube ends as shown in Fig. 206. 

[7] Blisters and Laminations. 

With modern boiler material these defects are happily rare. 
In former years, and especially with iron plates, they were only 
too frequently met with as referred to in Sec. 5. The chief dan- 
ger from these defects is due to the weakening of the plates and 
the liability of overheating, due to poorer conductivity for heat. 
Such defects if very small are often left undisturbed, with care- 
ful watching and measurement from time to time. If small and 
the metal quite thin on one side, the thinner part was cut away, 
leaving the thicker side to do duty for both. In some cases also 
a dog and supporting bolt was fitted to support the remaining 
metal. For some serious cases, however, it was usually con- 


sidered preferable to cut out the metal thus affected, and cover 
the hole with a patch. 

In all cases where patches are put on, or in general where 
new material is put into the boiler, it is well, if convenient, to 
select it of stock as nearly like the boiler as possible in physical 
and chemical constitution. This will tend to decrease the pos- 
sible sources of electro-chemical action as discussed in Sec. 40. 

[8] Tubes. 

The repairs to boiler tubes comprise re-expanding, plug- 
ging, and renewal. 

Expanding has been explained in Sec. 16, and a repetition 
of the process may be required from time to time, to keep the 
tube ends tight. The immediate cause of the leakage of boiler 
tubes is often the accumulation of dirt and scale about the tube 
ends and on the tube sheet. These points should therefore be 
carefully looked after, or the re-expanding will be of little use. 
Care must be taken that the operation of re-expanding is not 
repeated too often, or the metal of the tube end may become so 
thinned and hardened that there will be danger of weakness or 
brittleness at this point. 

For stopping a tube which has split or in any way de- 
veloped a serious leak, a tube stopper or plug is used. Tem- 
porary stoppers of pine wood are often employed. Such a plug 
closely fitting the tube and twelve or fifteen inches long may 
be forced in from the front end to a point where it covers a split 
or hole, and thus provides a temporary repair. The swelling of 
such a plug caused by the action of the water and steam will 
cause it to stick closely in place and thus more effectually stop 
the leak than would a metal plug in the same location. For more 
permanently plugging a tube, tapered cast iron plugs are used, 
one at each end, driven in to a tight fit and held in place by a 
rod passing through them from one end of the tube to the other 
and set up with thread and nut. The plugs and nuts should be 
faced so that when set up with a copper washer or turn of cop- 
per wire underneath, a steam tight joint between the plug and 
nut may be made. To plug a tube in this manner the fire 
must, of course, be drawn and the back connection entered in 
order to insert the back plug in place and adjust the rod and nut. 

When a tube is to be renewed the old one must first be 
drawn. To this end the back end is closed down so as to readilv 


pass through the hole in the tube sheet. A rod is then passed 
through the tube and a nut and washer are fitted to the back end, 
the washer being of such size that it will bear on the tube end 
and at the same time will pass through the hole in the tube 
sheet. The front end of the rod passes freely through a dog or 
strong-back whose feet rest on the tube sheet, and is provided 
with a nut bearing on the face of the dog. The nut being then 
forced down, the rod is withdrawn and with it the tube. To 
facilitate withdrawal the front end of the tube is usually made 
of slightly larger diameter than the back end, and the tube once 
started is readily withdrawn the remainder of the way. Before 
inserting the new tube the metal of the tube sheets about the 
holes should be carefully examined to see that the holes are 
smooth and fair and that no reason exists why the expansion of 
the new tube may not make a steam tight joint. The new tube 
may then be inserted, the ends expanded, beaded over at the 
back end or at the front and back ends if no stay tubes are 
fitted, and the operation is complete. 

[9] Leakage About Stays and Braces. 
Leakage at these points may be due to corrosion about the 
joint, or to loosening due to repeated expansion and contraction, 
or to bending or distortion of the plate or stay caused by 
bulging or partial collapse of the plate. If the leak is not serious, 
setting up on the nut, or caulking about the joint between stay 
and plate may prove sufficient. If not it will usually be neces- 
sary to remove and refit the stay. For screw-stay bolts this will 
usually require the reaming out and retapping of the holes for 
the next larger size. For braces fitted with nuts or nuts and' 
washers, the same size can usually be replaced, but especial 
care should be taken to insure the proper smoothness and fair- 
ness of surface about the holes, as well as the proper fit and ad- 
justment of the nuts and washers, so that the usual fitting will 
provide tightness of joint. 

[10] Bulging or Partial Collapse of Furnace or Combustion 
Chamber Plates. 

We have already discussed in Sec. 38 the causes of bulging 
or collapse, and the steps most suitable to insure immediate 
safety. When the time comes for examination and repair the 
following steps may be taken. 

If the bulge is quite small it may be decided to leave it un- 


disturbed, assuming that its strength is practically as great as 
before. In such case, however, a template should be fitted to the 
bulge, and this should be applied from time to time in order to 
detect any signs of further yielding at this point. 

In other cases special girder or through braces may be fitted 
to support the bulged portion, the details of the arrangement 
depending, of course, wholly on the circumstances of the injury. 

In other cases the bulged part may be more or less com- 
pletely forced back into place. This, however, is an operation 
requiring both skill and care, and should not be undertaken 
without making the preparations necessary to carry it out in 
the proper manner. 

A portable grate or furnace for burning coke or charcoal is 
first provided of such shape and with such arrangements that 
the burning fuel may be brought close up to the bulged surface. 
An artificial draft may be provided by means of a blower from 
a hand forge fitted with a suitable conduit, or otherwise as may 
be most convenient. In the meantime a cast-iron former block 
should be provided, shaped on its face to correspond to the sur- 
face of the plate when forced back into position. A hydraulic 
jack or other like appliance must next be provided, taking es- 
pecial care to arrange for the distribution of the load on the 
base over a considerable area of plate so that no harm may be 
done by the reaction from the head. To this end a support of 
heavy timbers is usually the most convenient to arrange. These 
various appliances being in readiness the bulged plate is heated 
to a low red, the former-block and jack are adjusted in position, 
and the plate is carefully forced back into shape. Several ap- 
plications of the heat and of the jack may become necessary 
before the plate is restored to its original form. 

It is thought by many that the partial heating of a steel 
plate in this way with no later opportunity for general annealing 
is liable to injure its homogeneity and toughness, especially if 
the operations are carried on at too low a temperature, or ap- 
proaching what is known as blue heat. This was undoubtedly 
true for much of the earlier products of the steel makers, and 
there is no doubt that the homogeneity of the metal is thus 
somewhat disturbed. With the latest and best grades of boiler 
plate, however, little danger as regards strength at least, need 
be feared on this score, and but little hesitation is now felt in 
reducing to shape a bulge of moderate depth. 


[n] Split in Feed-Pipe. 

A small split in the feed-pipe may sometimes be temporarily 
repaired by wrapping with heavy canvas and marline, or copper 
wire, it is, however, difficult to make a tight joint with hand 
wrapping, especially with modern high pressures, and a more 
effective plan is to form up a patch of sheet copper and secure 
it in place by bolted strap clamps, with a sheet of good elastic 
packing, or a thin layer of stiff putty between the patch and the 
pipe to make the joint. 

A small hole in the feed-pipe may, if the metal is thick 
enough, be stopped by drilling and tapping out and riveting 
in a small screw plug. Soft solder when applied with skill may 
also be used to stop pin-holes, or to aid in securing a suitable 
plug in a larger hole. If, however, the hole is of any consider- 
able size, or if the metal is thin about its edges, some form of 
patch, as described above, must be made use of. 

All such repairs are, of course, only temporary, and at the 
earliest convenient opportunity the damaged length of pipe 
should be replaced with new. 

In some cases with copper pipes, however, where the neces- 
sary materials and skill are available, it may be desired to under- 
take a more permanent repair by brazing a patch over the hole 
or split. To this end the metal about the defective spot is first 
cleaned with file, emery cloth and acid. The patch similarly 
cleaned is cut from sheet copper of about the same thickness 
as that of the pipe, formed to fit over the defective spot and 
wired in place. A clear coke or charcoal fire is then prepared 
on a forge and the pipe placed in position. The spelter or hard 
solder, mixed with borax as the flux, is then placed in position 
on the inside of the pipe and the whole carefully heated. At 
the proper temperature the spelter will melt and run in between 
the patch and pipe, thus forming a joint between the two. Care 
must be exercised in this operation in order to avoid danger of 
overheating or "burning" the copper. The resulting loss of 
strength has been referred to in Section 3. 


We shall undertake in this section only a few hints regard- 
ing the more common operations involved in the overhauling, 
adjustment and repair of marine machinery. 


[i] Cylinders. 

First, in the main engine, the cylinder covers must be re- 
moved at proper intervals, and attention given to the condition 
of the wearing surfaces and of the piston springs. The troubles 
most liable to be found are cutting and scoring of these sur- 
faces, and derangement or breakage of the springs. The nuts of 
the follower studs and all other forms of screwed fastenings 
should also be examined, in order that any tendency toward 
loosening up or backing off may be noted and checked. 

To examine the condition of the piston the follower plate 
is lifted and the springs and packing rings removed. The latter, 
if wearing properly, will be of uniform thickness around the en- 
tire circumference, and of uniform polish on the outer surface. 
If the piston-rod is bent or cylinder bore not quite in line with 
the motion of the rod, the ring will wear wedge-shaped in cross 
section. In replacing, care should be taken to note the degree 
of tightness of the ring when set out with its springs. This 
should be such as to support the ring anywhere along the bore, 
but not so much that the ring cannot be pushed along by hand, 
using moderate force. In closing up the cylinder or valve-chests, 
especial care should be taken to see that all nuts, split pins, etc., 
are properly secured in place, and that no tools, waste or other 
foreign substance are left within. Neglect of this latter point 
has often been the cause of serious trouble, resulting in 
broken cylinder-heads, bent piston-rods, broken valves, port 
metal, etc. 

To test the tightness of the joints of the cylinder liners, the 
cylinder being open, steam is admitted to the jacket and the 
joints are carefully examined top and bottom. Where there is 
no manhole on the lower head, as is common with small cylin- 
ders, a leak of any significance will become evident by opening 
the lower indicator cock. 

In order to test the tightness of the piston under steam, 
the cylinder-head being in place, we may proceed as follows : 
Put the engine and links in such position that a little steam can 
be admitted on top of the piston and then open the bottom 
indicator cock. Then admit the steam, usually through the 
starting valve, and if it blows through the lower cock the piston 
is thus shown to be leaky. For a thorough test it will be well 
to try the leakage in both directions; that is, from top to 
bottom, as above, and similarly, from bottom to top. 


[2] Pin Joints and Bearings. 

The various pin joints and cylindrical bearings will need 
attention according to the special circumstances of the case. The 
manner in which a joint or bearing has been working, both as 
to noise and temperature, will often serve as a guide to those 
which the most require attention. 

To examine the crosshead bearings the main points of the 
operation may be outlined as follows : 

(1) The crosshead with piston-rod and piston must be se- 
cured near or at the top of the stroke. This is sometimes done 
by inserting pins or bolts in the upper edge of the slide through 
the holes for securing the slipper, and allowing such to project 
over the top of the guide. Or otherwise it may be done by 
shoring in such way as may best be suited to the details of the 
case in hand. 

(2) The outer caps and brasses are removed. 

(3) A wooden chock lashed to the connecting rod is fitted 
between its upper end and the guide. This will serve to support 
the upper end when free from the crosshead. 

(4) The connecting rod sloping in such way as will bring the 
support of its upper end upon the chock, the turning engine is 
used to revolve the crank down until the parts are sufficiently 
clear to admit of the examination desired. 

To remove the crank-pin brasses the main points of the 
operation may be outlined as follows : 

(1) Starting with crank near its lower position the bottom 
cap is removed and landed on a bed suitably prepared for it. 

(2) The turning engine is then used to carry the crank up 
nearly or quite to the top of the stroke, where the crosshead is 
secured by hanging up or shoring as described above. 

(3) The upper brass is then secured and the crank is rotated 
down until the parts are sufficiently clear for the purpose in 

For like examinations of other parts, similar means will 
readily suggest themselves. 

For removing the lower brass of the main pillow-block 
bearings where it is in the form of a half cylindrical shell, it 
needs simply to be rotated out as noted already in Sec. 21 [ii]. 
A method sometimes used for this purpose is to clamp to the 
crank-arm a bar of steel carrying a projecting pin or bolt so 
placed that it will engage with the top face of the brass when 


rotated around. The upper cap and brass being then removed, 
the shaft is rotated carefully by means of the turning engine, 
and in this way the lower brass is forced around and up until 
free from its bed. 

Bearings and journals of this character may simply require 
readjustment, or refitting and adjustment as well. When the 
wear has been considerable, the liners or chock pieces between 
the brasses will need thinning, in order to reduce the clearance 
between the journal and the bearing to the propor amount. In 
order to measure the clearance under any given adjustment, a 
piece of lead wire is employed, of somewhat greater diameter 
than the clearance is to be. This wire is placed on the journal 
and the cap is screwed down hard, thus compressing it between 
the journal and the brass. The cap is then removed, and from 
the resulting thickness of the wire the clearance at any point 
between the journal and the brass may be readily measured. 
Such an operation is called taking a lead. Several leads, if de- 
sired, may be taken at once from a bearing, and will thus serve 
to map out quite satisfactorily the distribution of clearance, and 
thus to show when the proper adjustment has been made. The 
proper amount of clearance in any given case is somewhat a 
matter of judgment, and will, of course, vary with the size of the 
journal. In ordinary cases it may be made about .002 of the 

If time is limited the adjustment may be effected without 
the taking of leads, as follows : The chock pieces or liners are 
taken out and the nuts are tightened up till the brasses bear full 
on the journal. Their positions are then marked, and they are 
then backed off an amount determined by judgment or ex- 
perience with the particular circumstances of the case, and the 
liners are stripped to fit this adjustment. While this method is 
not as satisfactory as with leads, it is much quicker, and with 
experience good results may be obtained. 

In connection with the marking of the nuts for such pur- 
poses it may be recommended as a good plan to mark per- 
manently with a line and a numeral i, 2, 3, 4, 5, 6, the faces of 
all the large nuts likely to be used in making adjustments. An 
adjacent reference line on the bolt or on the metal of the bearing 
caps will then furnish means for making a record of each adjust- 
ment, or if no permanent record is desired, such me^ns will 
greatly facilitate making the adjustment in any given case. 


The refitting of bearings and journals on shipboard does 
not usually extend beyond removing or smoothing rough spots 
caused by overheating and scoring. This may be done by filing 
followed by "lapping" with oil stone powder or dressing with an 
oil stone. In some cases emery is used, but great care is then 
necessary in order to remove all particles, as if allowed to re- 
main they will give trouble by continued cutting and grinding 
in the bearings. The brasses are similarly redressed, usually by 
scraping. The nature of the contact between the brass and the 
journal is tested by lightly smearing the latter with red lead 
and then applying the brass in place and lightly rotating : 
and forth. The high spots will then be shown by the red lead 
and may be further dressed down till a satisfactory fit is obtained. 
It may be noted that where brasses are thus fitted up each 
one may show a satisfactory fit when tried separately and with- 
out external constraint, and yet when in place they may be 
unable to come to the same relative positions on the journal and 
make satisfactory contact with it. For this reason it is prefer- 
able to test the contact with the brasses together and regularly 
secured in place. This is not always possible, however, by reason 
of the additional time required, and judgment in all such cases 
must be used, having in view the various circumstances of the 
case in hand. 

[3] Cross Head Guides. 

The main guides and cross head slides must receive atten- 
tion, both as regards cutting or uneven wear, and as regards 
the question of adjustment. The result of wear is to throw a 
cross-breaking stress on the piston-rod at each stroke, as the 
cross head is forced over by the oblique action of the connecting 
rod to a bearing on the guides. Care must therefore be taken 
that when the engine is turned without a load, the guide surface 
remains in full contact with the face of the slide, and therefore 
in a condition to support and guide the lower end of the piston- 
rod in a path consistent with the movement of the rod in the 
axis of the cylinder. If this condition is not fulfilled the neces- 
sary adjustments must be made in the manner be*t suited to' 
the structural arrangements of the case in hand. 

To remove the slipper or bearing piece for examination or 
refitting, screw eyes may be screwed into holes in its upper 
edge made for the purpose, and from these the slipper may be 
supported by means of wire rope or stout wire wound around 


a bar suitably supported and secured above the slipper. The 
bolts holding the slipper to the crosshead are then removed and 
the crosshead forced over by screw or hydraulic jack or other 
convenient means, sufficient to ease the pressure between the 
slipper and the guide. The slipper may then be lowered by using 
the bar as an axle and the rope or wire will readily follow down 
between the crosshead and guide surface. 

[4] Crosshead Marks. 

In connection with the adjustment of the moving parts of 
the engine it is well to have a mark on the crosshead and cor- 
responding marks on the guide, showing the extreme positions 
of the piston when in contact with the cylinder heads, top and 
bottom; also two marks placed slightly within the latter and 
showing the ends of the natural stroke, and a mark placed mid- 
way between the two latter, showing the location of the piston 
when in midstroke. The distance from the extreme marks to 
those showing the ends of the stroke, shows the amount of 
clearance proper between the piston and the cylinder heads 
when the former is at the ends of the stroke. 

This will vary with the size of the engine and character of 
the workmanship, but it is usually found between y and ^ or 
24 inches, being a little more at the bottom than at the top, to 
allow for the general tendency of the parts to lower rather than 
to rise through the effect of wear. 

To determine these points a convenient reference mark is 
first placed on the crosshead. The connecting rod may then be 
disconnected and the parts hoisted up as far as they will go, 
or until there is contact between piston and head. A mark is 
then made on the guide corresponding to that on the crosshead. 
The parts are then lowered down as far as they will go, or until 
there is contact between the piston and the lower head, and 
anothef mark is made on the guide corresponding to that on 
the crosshead. The distance between these is then taken, and 
from it is subtracted the length of stroke. The remainder is 
then divided between the two clearances, top and bottom. 
Midway between the two inner points a point may be placed 
to indicate the location for mid or half stroke. 

Thus, if the stroke is 36 inches and the distance found as 
above is 37 inches, the I inch difference is to be divided between 
the two clearances, giving to the upper, say, 7-16, and to the 


lower 9-16 inch. These differences are then laid off within the 
outside marks, and the points thus given will serve at any time 
as a guide for the adjustment regarding clearance proper, while 
the movement of the piston may be readily brought to conform 
to these limits by suitable adjustment of the liners or chock 
pieces in the joints and bearings of the connecting rod and 
crank-shaft. The 36 inches may then be divided equally and 
the mark placed to show mid stroke, such a point being some- 
times of use in connection with the setting of the valve. 

Another method of determining the clearance which is avail- 
able when the cylinders have manholes is as follows : The man- 
holes are removed and a number of balls of stiff red lead or other 
putty and faced with plumbago are distributed on the top of the 
piston and on the inside of the lower head. The engine is then 
given a revolution by means of the turning engine and the 
balls are collected. This method serves to show just how the 
clearance is distributed, and is therefore a valuable test for a 
bent piston-rod, a condition which will throw the piston out of 
position and give greater clearance on one side than on the 

[5] I/ining Up. 

An important feature, both of the original setting up of ma- 
chinery and of its overhauling and adjustment, is the determina- 
tion or correction of the alignment of the various moving parts. 
It is clear, of course, what the condition of proper alignment is. 
For a marine engine it may be briefly stated as follows : 

(1) The centers or axes of the main pillow block bearings 
should all be in one straight line, which will coincide with the 
crank-shaft axis, and which we will take as a standard or line of 

(2) The axes or center lines of the cylinders should all be 
in the same vertical plane containing the center line of (i), and 
they must also be at right angles to this line. 

(3) The same vertical plane should also contain the axes or 
center lines of all the crosshead pins. 

(4) The axes or center lines of the crank-pins or crank-pin 
bearings in the connecting rods must also be parallel to the 
center line in (i). 

(fO The surfaces of the main guides must be parallel to the 
plane in (2). 

(6) The line shaft must be in line with itself, and unless a 


flexible coupling is provided between this and the crank-shaft 
it must also be in line with the crank-shaft, as determined by the 
line in (i). 

The same general principles, of course, control the align- 
ment of the valve gear and the various other moving parts, such 
as the starting, handling and drain gear, attached pumps, etc., 
but into these we need not go in further detail. 

The implements used in establishing the relation between 
these various lines and planes will naturally vary with the circum- 
stances, but are usually found among the following: 

The level, plumb line and square. 

The straight edge. 

The stretched wire or cord. 

Provision for using a line of sight. 

A straight line, such as that for a line of shafting, may be 
determined by either of the latter. The sag in a piano wire of 
known size and length and stretched with a known weight over 
a pulley is a matter which may be found by a computation, into 
which we cannot enter here, or, better still, it may be deter- 
mined in the open air by the aid of a surveying instrument with 
the usual cross hairs. A table giving the sag at various points 
between the ends for known lengths and for a known stretching 
weight will then give a leady means of establishing a straight 
line on board ship, by stretching the wire under the same con- 
ditions, and then setting upward at the various points the amount 
of the sag. A series of levels may thus be found which will give 
a true, straight line within the limit of the error in measure- 

When a line of sight is used the following method may be 
employed : A board is fitted to the bearings at the extreme ends 
of the line to be run, a hole of some considerable size being 
made in the board at about the center. This hole is then covered 
with a piece of thin sheet metal having a small hole, say, 1-16 to 
% inch in diameter. These sheets of metal are adjusted by 
measurement until the small hole, as accurately as may be, is 
brought to the center of the bearing. Similar boards are then 
prepared for the other bearings or points at which it is desired 
to establish the line. A light is then placed at one end beyond 
the further hole, and the eye at the other end. An assistant then 
adjusts the intermediate pieces of sheet metal until the light 
reaches the eye through the entire series of holes. The centers 


of these holes will then serve to establish a series of levels which 
may be marked on the pedestals, bulkheads or other convenient 
points, and which will serve to establish the line as desired. 

In lining up and adjusting the engine itself so that the 
various conditions (i)-(5) above are fulfilled, very much will 
depend on the accuracy and care with which the various pans 
have been machined in the shops. The gaps in the bed plate, 
which receive the main bearing boxes, should be planed out so 
that they are all in line. This is a matter which may be tested 
by a straight edge, or by measurements from a stretched wire. 
If the bottom brasses are then adjusted, all to the same thick- 
ness, they will evidently support the crank-shaft in line: This 
may also be further tested by reference to a line of levels, each of 
which is obtained by measuring the same distance upward from 
the bottom of each gap, and locating on the bed plate at any 
convenient point the level thus determined. This adjustment, 
it may be noted, will bring the axis of the crank-shaft parallel 
with the bottom of the gaps. 

Passing now to the columns, the seatings for their feet 
should have been planed at the same time as the gaps. This will 
bring the bottom of the feet in a plane parallel with the bottom 
of the gaps, and hence parallel with the axis of the crank-shaft. 
The columns and cylinders may now be erected and temporarily 
secured in place. The center line for each set of moving parts 
may be next determined as follows : A piece of board is placed 
across the top of the cylinder and secured at each end and ovei 
one or more stud bolts. Then, by caretul measurement or the 
use of a beam compass, the center of the bore at the top of the 
cylinder is located on the board. A small hole is then bored 
at this point and a fine wire is passed through and attached to 
a little frame, which will allow a few inches of the wire to show 
above the board. The lower end of the wire is then located in 
the crank-shaft axis by suitable measurement from the faces of 
the bed plate. The wire is then stretched between the two 
points, and the adjustment of the upper end verified by renewed 
measurement. If these parts come accurately to place we shall 
find that the wire will be truly central relative to the opening 
left in the lower head for the stuffing box, and hence it may be 
taken as the true center line for the cylinder as a whole. We 
shall also find the wire at ,-i constant distance from the guid^ sur- 
face on the column, thus proving the latter parallel to the center 


line, as required by (5) above. We shall also find the wire at 
right angles to the crank-shaft axis, as required by (2) above, 
and, of course, in a general way, at right angles to the bed plate 

Supposing these various conditions fulfilled for each center 
line separately, then we must see if they are all in the same 
fore and aft plane as referred to in (2) above. To examine this 
point we may try to look the wires "out of wind," as the ex- 
pression is, by standing aft or forward and trying, by sighting 
fore and aft, to bring all the wires accurately behind the nearest 
one. If these various conditions, for each wire separately and 
for all collectively, are not fulfilled, then the necessary adjust- 
ments must be made until they are brought into the required 
relations, as above specified. As a further adjustment to be 
made at this time, a wire may be stretched fore and aft along or 
near the guide surfaces, and at the same horizontal distance 
from each vertical center line. This will serve to show whether 
the guide surfaces stand in the right direction fore and aft, and 
hence, whether the plane of each is parallel to the central plane 
defined in (2) above. 

In setting up attached auxiliaries, such as air, circulating 
or feed-pumps, the principles used will be similar to those al- 
ready discussed, while the methods used will depend upon the 
particular circumstances of the case. If the seatings are located 
on the bed plate they will probably be planed at the same time 
as the main bearing gaps, and if the pump bases are faced at the 
same time they are bored, their axes will come at right angles to 
the seatings, and hence parallel to the axes of the main cylinders, 
and thus properly in line. 

In order to test the line of the crosshead pins the connecting 
rod may be hung from the upper end and allowed to swing 
partially free at the bottom. Then, by turning the engine into 
various positions in the revolution and measuring the fore and 
aft clearance between the crank webs and the faces of the lower 
end of the rod when thus relieved of constraint, the accuracy of 
the adjustment can be determined. This operation may be car- 
ried out in the following manner: Assuming that it is desired 
to test the adjustment in the case of an engine completely set 
tip, the cap on the lower end of the connecting rod is removed 
and the cross head is shored up in any convenient manner. The 
engine is then turned slightly by hydraulic jack or turning en- 


gine in such direction as to carry the crank-pin away from the 
rod and thus leave the latter free in a fore and aft direction, so 
far as direct contact with the pin is concerned. Measurements 
are then taken, and the engine is next moved around so as to 
bring" the pin against its bearing again, and then on to a new 
position. Here the crosshead is again shored, the pin backed off 
measurements taken as before, and so on. In this way the piston, 
crosshead and connecting rod are carried around by resting on 
the crank-pin, and then in each position the connecting rod is 
freed by shoring the crosshead and backing off the crank-pin. 
In case the clearance between the crank-pin brass and the faces 
of the webs is too small to allow a possible irregularity of move- 
ments to show itself, the regular brass may be taken out and a 
dummy brass or wooden block, with sufficient clearance for all 
possible motion, may be fitted in its place. 

In locating a line shaft, a wire or line of sight may be used 
in the general manner above described. After locating the center 
of the stern tube at the stern post, the line is run to the after 
end of the crank-shaft or on through to the forward end, as may 
be desired, such point being located by measurement accord- 
ing to the drawings of the ship, engine seating and bed plate. 

In twin screw ships the centers at the after struts are located 
by measuring at the proper height the necessary distance out- 
ward from the center line of the stern post, at the same time 
squaring from this center line so as to bring both centers at the 
same height above the keel. The two lines are then set at the 
forward end at equal distances from the keel at the proper height, 
in accordance with the drawings. In most cases the shafts are 
not parallel with the keel, either in a vertical or horizontal direc- 
tion, usually inclining outward from forward aft and either up- 
ward or downward, according to the size of the ship and other 
circumstances of the case. 

In the examination of the adjustment of machinery which 
has been already in operation, as in the routine care of marine 
engines, it is not to be supposed that the preceding operations 
in all their details will be necessary. Judgment must be used, 
having in view the particular points to be examined, and the 
best means of effecting such examination in acordance with the 
general principles above discussed. 

Thus the fairness of the line shafting must be tested occa- 
sionally, lest, as time goes on, wear in the bearings or changes in 


the structure of the ship may throw it seriously out of line. For 
this purpose the following method may be used : Three or more 
laths or battens are provided with a straight edge on one side. 
These are then placed across the shaft on the upper side, as 
nearly horizontal as judgment may indicate, being located at 
successive points along the line of shafting to be examined. A 
line of sight is then taken along the projecting lower edges of 
the battens and by adjustment they are brought parallel to each 
other. Then, if they can all be brought into one line the shaft 
itself is in line. If not, the shaft is out of line in the vertical 
direction, and, by blocking or other similar adjustment, until the 
battens are all brought into one line the amount by which the 
shaft is out at one point relative to two others is readily de- 

In a similar manner the line in the horizontal direction may 
be tested by holding the battens vertical against the side of the 
shafting. This method presupposes, of course, that the shafting 
throughout the part tested is all of the same size. 

Instead of placing the battens on the shafting they may be 
placed on the couplings quite as well, supposing, of course, that 
the latter are all of the same size. 

Another test which is sometimes used consists in slacking 
the coupling bolts at a coupling near which it is suspected that 
the shaft is not in line, and noting whether there is any tendency 
for the two parts of the coupling to open out on one side or 
another. Such an opening out then shows an unfairness in the 
line in one way or another, according to the side on which it 

[6] Valve Gear. 

In the routine examination of the valve gear the points to 
be looked after are wear and lost motion in the excentrics and 
excentric straps, in the link block brasses of the Stephenson link, 
or in like parts of other forms of valve gear, in the various pin 
joints and connections, and abrasion or uneven wear in the 
valve faces or seats. In the case of the excentrics and straps and 
various pin joints, the wear may be taken up by the adjustments 
usually provided and in the general manner already outlined for 
other similar parts. Excessive or irregular wear in a flat slide 
valve or seat may require either resurfacing or renewal, accord- 
ing to the circumstances of the case. With piston valves, es- 
pecially if not fitted with rings, excessive wear in either valve 


or seat will mean a serious steam leak, and will usually call 
for new parts, either for valve or seat or both. 

[7] Thrust Bearings. 

The most common trouble with thrust bearings is scoring 
or irregular wear of the bearing surfaces, or a general wear 
which may allow the thrust shaft to move forward sufficiently 
to put a pronounced thrust on the line and crank-shafts. This 
will result in heating and wear, and may throw the crank-shaft 
bearings out of line. A little clearance is usually allowed in the 
shaft couplings and in the fitting-up of the crank-shaft and crank- 
pin bearings, so as to insure freedom from end thrust for the 
crank-shaft as a whole. If the wear at the thrust collars should 
exceed this, then a part of the thrust would be transmitted to 
the crank-shaft, with effects as above noted. 

With adjustable or horse-shoe collars this trouble is readily 
adjusted by moving the collars back by means of the adjusting 
nuts provided. With the plain type of thrust bearing as in Fig. 
149 the bearing as a whole must be moved slightly aft by means 
of the screws provided for the purpose. 

[8] Circulating Pump. 

The engines for operating such pumps require the care 
necessary in all machinery of such type, the principles of which 
have been already discussed. The runner simply needs examina- 
tion from time to time to make sure that it is running freely, 
but without undue clearance on either side in its casing. Trouble 
is often met with in the maintenance of circulating pumps by 
the choking of the inlet passage, valve or strainer by marine 
growth of one form or another. To aid in freeing the strainer 
and inlet passage a connection is often made with the delivery 
of one of the fire or other auxiliary pumps by means of which 
they may often be cleared without the need of regular over- 

[9] Condensers. 

The chief troubles to be expected with condensers are as 
follows : 

(1) Leakage about the tube ends through the packings from 
the salt water to the steam side. 

(2) Corrosion and pitting of the tubes, resulting ultimately 
in the development of holes or the breaking of the tubes and 
similar leakage, as in (i). 


(3) Fouling of the tubes with oil deposit on the condensing 
surface, thus decreasing the heat conductivity of the metal and 
the efficiency of condensation. 

The condenser heads or bonnets must therefore be taken off 
from time to time and the condition of the tube ends and pack- 
ings examined with reference to the matter of leakage and 
general condition. 

To test the condenser for leakage the main inlet and out- 
board valves should be tightly closed, and the connections to 
the low pressure cylinder and air pump closed by blank flanges. 
The condenser heads may then be taken off and the steam side 
filled with water while a watch is kept for leaks on the tube 
sheets at each end. It is also desirable to be able to put, by 
means of a hand-pump or otherwise, a pressure of 15 or 20 
pounds per square inch on the contained water, thus making 
the development of the leak more certain. It is sometimes 
desirable to be able to identify both ends of the same tube. This 
may be readily done by passing a wire through from one end to 
the other. In some cases a lamp held at one end will serve the 
same purpose. 

When provided with bonnets for the purpose, the steam 
side of the condenser must also be occasionally opened and the 
tubes and interior surfaces examined, with reference to grease 
coating and general condition. Where there may be any doubt 
as to the condition of the tubes in the interior, a few should be 
drawn when the heads or head bonnets are removed, and their 
condition determined. 

So far as the accumulation of grease is concerned, the con- 
denser may be cleaned by the use of hot soda or lye water, care 
being taken to wash it out thoroughly so as to remove any 
excess of the alkali. When soda is used for the boilers it is some- 
times introduced into the condenser, there entering the feed 
water and then passing on to the boiler. It may be questioned, 
however, whether this is the best plan, as accumulations of 
grease only partly converted to soap may thus be carried into the 
boiler, there giving trouble, as already referred to under that 

Zinc plates are often used on the salt water side in condenser 
heads to protect against corrosion, in a similar manner as ex- 
plained for boilers in Sec. 40. The condition of these plates and 
of their attachment to the shell should be carefully noted when 


the condenser is opened, and such repairs made as may be 

[loj Air Pumps. 

The head, foot and bucket valves require the most frequent 
and careful examination. They often tend to become coated by 
an accumulation of a black, greasy paste formed from the cylin- 
der oil and the material of the wasted zinc plates, if such are 
used. This accumulation may prevent their proper working, and 
they should be carefully cleaned as occasion may offer. With 
air pumps attached to the main engine, and where the number of 
strokes is usually greater than with the independent pump, the 
valves sometimes give trouble by severe pounding against their 
seats and guards. This is due to their inertia, and is likely to 
be more severe as the valves are heavier and have more litt. 
Light metal valves of sufficient number and size to allow of 
moderate lift are therefore to be preferred for all such purposes. 
(See Figs. 185, 186). 

[n] Pumps in General. 

The chief troubles to be expected in the operation of the 
various forms of independent pumps found on shipboard are 
as follows : 

(1) In the water end the plunger rings or packing or the 
barrel may become worn, thus allowing considerable leakage or 
"slip," especially under high pressure, as in boiler feed pumps. 

(2) The water valves, either inflow or delivery, may become 
worn or deranged, and thus fail to hold the water as they should, 
so preventing the proper operation of the pump. 

(3) In the steam end the piston rings or cylinder barrel may 
be.-ome worn, allowing steam to blow from one side to the other 
and decreasing the effective steam load on the piston. 

(4) The main steam valve or more often some part of the 
auxiliary steam operating gear may become worn or deranged 
so that the pump can no longer properly operate under steam. 

These various troubles must be guarded against by periodi- 
cal examination of the points mentioned with a view to wear or 
derangement of any kind whatever. 

[12] Piping. 

Steam piping of copper if properly constrained may become 
brittle and weakened in spots by long continued expansion and 
contraction. An indication of this mav often be found in the 


wavy or irregular condition of the surface. Such pipe should, 
of course, be replaced at the first opportunity. A repair may, 
however, be made by banding with screw clamps made of strap 
iro-i or steel and closely spaced over the suspected part. 

Small holes which may sometimes develop may be treated 
with a soft patch held in place by a screw clamp, by filling with 
solder, or by a patch as in Sec. 42 [2] (10), according to the cir- 
cumstances of the case. Leakage and like trouble with steel 
piping may be treated by the same general means as for boilers, 
and as discussed in the preceding section. 

Sec. 44- SPARE PARTS. 

In order to provide for the results of regular wear, and for 
the possibilities of accident it is customary to carry a certain 
number of spare parts, especially of those most likely to require 
replacement either as the result of wear or accident. The pieces 
carried and their number will depend entirely of course on the 
extent to which it may be necessary or desirable to fit out the 
ship with provision for such wear and emergency. No attempt 
will therefore be made to give any complete list of such parts, 
but among those more commonly carried the following may be 
mentioned : Grate bars, bearers and dead-plates, furnace and ash- 
pit doors, boiler tubes, manhole and handhole plates with fit- 
tings, safety valve springs, boiler gauge cocks, fittings for boiler 
water gauges, feed check valves, bottom blow valve, surface 
blow valve, piston and pump rods for the various pumps, valves, 
valve guards and springs for the various pumps, follower bolts, 
nuts and springs for the various steam pistons, brasses for the 
various bearings, horseshoes for the thrust bearing, propeller 
blades, valve stems, metallic packing for the various stuffing 
"boxes where it is used, shaft coupling bolts, emergency shaft 
coupling, condenser tubes and glands, evaporator and distiller 
tubes, evaporator coils, one section of crank-shaft, if made in 


The chief dangers to marine machinery when not in use arise 
from the likelihood of rust and corrosion. Fundamental prin- 
ciples relating to these chemical changes have been discussed in 
Sec. 40, and by reference to that point it will be seen that the 
chief points to be attended to relate to the protection of the sur- 
face from moisture and corroding acids. Where applicable a 


good metallic paint well laid on will be found the most efficient 
and satisfactory. In such case the surfaces must be dry and 
well cleaned in accordance with the principles discussed in Sec. 42 
[i] (10). For finished surfaces or bright work generally, where 
paint would not be suitable, a coating of heavy cylinder or other 
like oil may be used, or perhaps even more commonly a mixture 
of white lead and tallow in about equal proportions. Either of 
these will efficiently protect the surfaces and will remain for a 
long period of time without becoming too hard to admit of ready 
removal, especially with the aid of a little kerosene or other 
light oil. 

One of the most important features connected with the lay- 
ing up of marine machinery is the getting rid of water contained 
in the various cylinders, pipes, bends, valve chambers, etc. The 
draining off of the water is, of course, of importance relative to 
the question of rust and corrosion, but it may be of even still 
greater importance relative to the question of freezing and 
possible rupture of the chamber, casing, or pipe containing the 
water. Many a cracked cylinder or valve chest or globe valve 
chamber, or split in pipe elbow or bend, or in boiler or condenser 
tube, etc., has been due to incomplete drainage of water and 
subsequent freezing. In laying up marine machinery, therefore, 
where there is any liability of freezing, a systematic study must 
be made of the piping systems, pockets, etc., where water might 
collect and by freezing result in damage. These remarks apply 
especially to piping and fittings, to small auxiliaries, to the con- 
denser, and to water-tube boilers. If the proper drains are not 
fitted and the water cannot be gotten out in any other way, then 
the necessary joints should be broken and the water removed 
in this manner. 




Sec. 46. SI/IDB VAI/7ES. 

[ij Simple Slide Valve. 

In Fig. 207 VW represents a simple slide valve, supposed 
to be surrounded by live steam in the steam chest C C. P and Q 


Fig. 207 Plain Slide Valve, Mid Position. 

are the^ ports leading to opposite ends of the cylinder as shown, 
while E is the exhaust port or passage leading to the condenser 
or to the air, as the engine is condensing or non-condensing. 
It is the business of the valve, as we shall explain later, to move 

Fig. 208. Plain Slide Valve, Position for End of Stroke. 

back and forth, thus alternately uncovering the ports P and Q 
and admitting steam from the chest to the ends of the cylinder. 
While the steam is thus being admitted at one end of the cylin- 
der, it must be allowed to escape from the other to the exhaust 



passage E, and thus the piston is moved to and fro in the cylin- 
der and the operation of the engine becomes continuous. 

If we suppose in Fig. 
207 that the valve is in the 
middle of its travel back 
and forth, or in mid posi- 
tion as it is commonly 
called, then the distance 
A B by which the edge of 
the valve on the steam 
side extends over the 
edge of the port is called 
the steam lap. Similarly 
the distance CD by which 
the edge of the valve ex- 
tends over the port on the 
exhaust side is called the 
exhaust lap. In Fig. 208 
let the piston be at the 
end of the stroke and the 
valve in the position 
shown. Then the distance 
A B by which the port is 
uncovered to steam is 
called the steam lead, while 
the distance CD by which 
the other port is uncov- 
ered for exhaust is called 
the exhaust lead. In gen- 
eral, therefore, lead is the 
amount by which the valve 
is open when the piston is 
at the end of the stroke. 
In Fig. 209 are 
shown a series of corre- 
sponding positions for 
valve and piston during a 
single stroke of the latter. 
The position a is the same 
as that of Fig. 208 and 
the port is open for the 


.admission of steam on I he left and for exhaust on the right. 
The piston is at the end of the stroke and just about to 
begin the stroke to the right. In b the piston has advanced 
about 10 per cent, of the stroke, and the valve has moved so 
that the port is nearly wide open. Up to this point the piston 
and valve have been moving in the same direction. In c the piston 
has advanced to about 60 per cent, of the stroke, while the valve 
has come back and has just closed the port to the entrance of 
steam. This is called the point of cut-off. Between b and c the 
piston has been moving on, but the valve has been moving back 
in the opposite direction. In d the piston has moved still further 
on while the valve is still moving to the left, and has just reached 
the point where the exhaust opening on the right is closed. This 
is called the point of exhaust closure, and the operation of com- 
pressing the steam in the end of the cylinder from this point to 
the end of the stroke is called compression or cushion. In e the 
piston is still nearer the end of the stroke on the right, while the 
valve has moved farther to the left, and is just about to open the 
port on the left for exhaust, thus allowing the escape of the steam 
which entered during the early part of the stroke. In f the piston 
has nearly reached the end of the stroke. The exhaust opening on 
the left is open still wider, while the port on the right is just 
about to open for steam. In g the piston is at the end of the 
stroke, the valve has moved so as to make the steam and exhaust 
openings still wider, and the return stroke is about to begin. 
This completes the history of the stroke, and the next or return 
stroke follows after it with like series of events, and so on con- 

[2] Double Ported Slide Valve. 

The valve shown in the above diagrams is called single ported 
because it covers but one set of ports or openings. A double ported. 
valve is shown in Fig. 210. This is a form of valve within a valve. 
Thus taking one end of the valve we have at AB one set of edges 
respectively for steam and exhaust, and at A! Bj another like set. 
Steam surrounds the outside of the valve and is therefore ready 
to enter the port P past the edge A when the valve moves suffi- 
ciently to the left. Steam likewise enters freely at the side into 
the passage S, and is therefore ready to enter the port P l past 
the edge A Jt as the valve moves to the left. Similarly as the 
valve moves to the right the ports P and P! are open to exhaust 
past the edges B and B 1? respectively. The passage E t leads 



over the transverse passage S, and thus the entire exhaust finds 
its way into E the outlet passage. 

With a double ported valve the area of the port opening 
required may be obtained with a travel of valve only one-half that 
for a single ported valve, or with the same valve travel, twice 
the area of port opening may be obtained. It is this feature 
which often leads to the use of a double ported valve where it 
is desired to obtain a relatively large opening with small travel 
of valve. 

[3] Piston Valve. 

The face of the valves in the types so far noted is a plane, or 
in other words, they are flat slide valves. If now we can imagine 
such a valve wrapped up so as to form a cylinder with the valve 

? B 


Fig. 210. Double Ported Valve. 

stem for axis, we shall have a cylindrical or piston valve as shown 
in Figs. 211, 214, 215. This consists essentially of two heads 
connected by an intermediate body, as shown. The steam enters 
past the outside edges, for example, and exhausts past the inside 
edges to the exhaust passage, in a manner entirely similar to 
that for the plain slide valve as above described. Otherwise the 
steam may enter past the inside edges and exhaust past the out- 
side as described below for the inside valve. The steam port and 
passage consists of an annular channel surrounding the valve and 
connecting with a passage leading to the end of the cylinder in 
the manner shown. The valve is placed somewhat out of center 
with reference to this annular passage, so that it is quite shallow 
on the side opposite the cylinder, and gradually increases in 
depth toward the side nearest the cylinder. This arrangement, 



which is shown in Fig. 212, gives a cross-sectional area of pas- 
sage varying in proportion to the amount of steam flowing 
through it, as may be seen by noting in the figure the natural 
direction of flow of the steam, radially outward through the 

Marine Engineering' 


Fig 211. Piston Valve. 

port opening, and then curving around to flow toward the 

Comparing the two forms of valve it is readily seen that in 
the piston valve the outer circumference represents the active 



part of the valve, and corresponds to the plate AB of the flat 
slide, as shown in Fig. 210. 

The great advantage of the piston valve lies in the fact that 
it is perfectly balanced as regards the steam pressures which 
act upon it. It is readily seen that the flat slide valve is forced 
against its seat by the excess of the pressure on its back over 
that on its face, which excess will in the usual case be large, 
and will give rise to a heavy frictional load to be overcome by the 
excentric acting through the valve stem. The piston valve, on 
the contrary, is forced equally in all directions, and hence moves 

Fig. 212. Section Through Valve Chest and Cylinder. 

freely so far as the steam forces are concerned, and with only 
such frictional resistance as may be necessary to insure tight- 
ness against steam leaks. 

In order to keep the heads of the valve steam tight they 
have been very commonly provided with one or more packing 
rings of similar character to those used on the main piston, but 
usually without auxiliary steel springs to force them outward. 
Such an arrangement is shown in Fig. 215. In the latest prac- 
tice, however, especially for quick-moving engines, the special 
rings are very commonly omitted entirely, dependence being 


placed on a good working fit at the start and on the rapid re- 
versals of motion, to reduce the leakage to a negligible amount. 
Such arrangements are shown in Figs. 211 and 214. In the latter 
a solid working ring is fitted as shown, in such manner that it 
may be readily removed and replaced with a new one as occasion 
may require. 

The valve seat is usually a separate piece of hard and fine 
grained cast-iron, fitted as shown in Fig. 211. The ports in this 
seat instead of being continuous all around, thus dividing the seat 
into separate parts, are usually bridged over at several points 
distributed about the circle. The head of the valve is thus car- 
ried across from one side to the other, and is prevented from 
catching or jamming, as would very likely occur without such 
bridging. Where valve packing rings are used it is especially 
necessary to provide such bridging in order to prevent the ring 
from springing out into the port opening, and thus jamming the 
valve, or causing other damage. In Fig. 213 is seen the de- 
velopment or lay-out of the port for the valve shown in Fig. 215. 

J/ai*in< Engineering 

Fig. 213. Development of Piston Valve Port. 

The bridges are placed on the slant so as to distribute as much 
as possible the wear on the valve rings. 

In the case of wear on the valve seat or rings, they may be 
replaced with new. When rings are not used the wear is 
usually less rapid, thus furnishing a further reason for their 
omission. The valve itself, however, will slowly wear, and as 
necessity may require a new head or new valve entire may be 

In some cases where a piston valve takes steam on the out- 
side, it is desired to lead the steam to the chest at one end only, 
and then to pass it down to the other end through the inside of 
the valve. In such case as shown in Fig. 215 the body of the 
valve is made hollow and as large as possible, thus connecting 
the steam chest at top and bottom as desired. 

The valve stem passes through the center of the valve and 
is usually secured with nuts at top and bottom, as shown in the 
figures. In case the valve is hollow for the passage of steam 



between the two ends, the stem must be carried in bosses sup- 
ported by radial arms connected with the valve heads, and as 
small as possible in order to present the least resistance to the 
flow in either direction. 

Figs. 214, 215. Piston Valves. 

[4] Equilibrium Piston. 

The work of moving the valve up and down which is thrown 
on the excentric may be much decreased by fitting an equilibrium 
piston as shown in Fig. 215 The cylinder in which this is fitted 
is open at the bottom to the valve chest, and hence the full 

3 66 


pressure of the chest acts constantly on the lower side of the 
piston. This may be so proportioned as to carry practically 
about all the direct weight of the valve, which thus floats on the 
steam, requiring comparatively small effort on the part of the 
excentric. to move it back and forth. 

Fig. 216. Joy Assistant Cylinder. 

[5] Joy's Assistant Cylinder. 

A further development of the equilibrium piston is found in 
Joy's assistant cylinder, as shown in Fig. 216. The purpose of 
this fitting is to more perfectly carry the weight of the valve uric! 
relieve the excentric and valve gear of the work required to 



move the valve. The cylinder is provided with steam and ex- 
haust ports, and in fact with its piston forms a small steam en- 
gine. The form of the piston as shown is such that it forms its 
own valve, thus simplifying the number of parts required. Power 
for assisting the main valve gear is thus obtained by the ad- 
mission of live steam to the assistant cylinder, and the ports are 
so arranged that cushioning at the ends of the stroke in this 
cylinder absorbs the forces due to inertia. An important ad- 
vantage claimed for such types of assistant cylinder over the 
ordinary equilibrium piston as in Fig. 215 is a considerable de- 
crease in weight, the greater efficiency of the apparatus for the 
purpose in view allowing of the use of smaller sizes. 

Such forms of equilibrium piston may, of course, be fitted 
to advantage with either the flat slide or piston forms of valve. 

[6] Equilibrium Rings. 

With a flat slide valve as in Fig. 210 the full steam chest 
pressure acts constantly on the back of the valve, while a much 
decreased pressure will act on a part only of the other side. In 

Fig. 211. Section Through Equilibrium Ring. 

consequence, the valve is forced with strong pressure against 
the seat and the frictional force thus developed must be over- 
come by the excentric. (See Sec. 73). With large valves and 
where the lubrication is scanty this may become excessive, 


causing the valve and seat to cut, and throwing a great deal 
of unnecessary work on the excentric. To relieve this condition, 
equilibrium or balance rings are often fitted on the back of the 
valve. Such an arrangement is shown in Fig. 217. The inner 
face of the valve chest carries a ring of metal within which is 
cut a groove as shown. Within this groove is a ring which is 
forced against the back of the valve by springs and screw ad- 
justment as shown. The back of the valve is faced off and thus 
a joint is made between the two, while the space within the ring 
and between the back of the valve and face of the cover is shut 
off from the steam within the valve chest, and hence this part 
of the valve is relieved from the pressure of the steam. In addi- 
tion to this, the space is sometimes connected by piping to the 
steam side of the condenser, thus bringing on this part of the 
back of the valve only the pressure in the condenser. In this 
way the load on the back of the valve and the resultant 
load on the excentric may be much decreased, and the opera- 
tion of the valves will be correspondingly smoother, and 
less work will be thrown on the excentrics and valve gear in 

There are several methods varying in detail for the fitting of 
the ring in such an arrangement, and in the formation of the 
joint between the ring and valve chest cover, but the principle 
is the same in all, and is sufficiently illustrated by the arrange- 
ment of Fig. 217. 

[7] Outside and Inside Valves. 

In the preceding figures for valves the steam enters past 
the outside edges and exhausts past the inside edges. Such is 
known as an outside valve. In some cases, However, it is con- 

Fig. 218. Inside Valve. 

venient to have this relation reversed, and to take steam past 
the inside edges, and exhaust past the outside edges. Such a 
valve is shown in Fig. 218, and is known as an inside valve. 

Live steam fills the space A and enters the port P past the 


edge C, while exhaust occurs past the edge F. The valve as here 
shown is in midposition and therefore C D is the steam lap and 
E F the exhaust lap. The only difference in the two forms of 
valve is in the relative amounts of outside and inside lap. In each 
case, for reasons which will appear later, it is seen that the steam 
lap is greater than the exhaust, being in one case on the outside 
and in the other case on the inside. It is also clear that the 
outside valve moves with the piston at the beginning of the 
stroke, and opposite to it during the latter part, while with an 
inside valve it moves opposite to the piston at the beginning, 
and with it during the latter part of the stroke. 


[i] Simple Excentric. 

We must now inquire by what means the valve can be 
given the motion necessary for the proper distribution of the 
steam as above described. The simplest of such means is the 
plain excentric as shown in Fig. 219. This consists of a cir- 
cular disc with center A set on the shaft excentric or out of the 
center. The distance between the two centers is seen to be OA. 
This is called the eccentricity or throw of the excentric." About 
the excentric is a strap ST, and attached to this is a rod RR. 


Fig. 219. Plain Excentric, Skeleton of Motion. 

As the shaft turns, the excentric turns with it, and thus gives 
to the rod a to and fro movement exactly as though it were a 
connecting rod attached to a crank OA. This principle of the 
equivalence between an excentric and a crank of equal throw 
is very important, and the motion should be studied until it is 
quite clear that the operation of an excentric is exactly the same 
as that of a small crank of throw equal to that of the excentric. or 
to the distance from the center of the excentric to the center of 

*Some writers understand by tlie term throw * twice the above distance. 
In the present work we understand the term to refer to the distance OA as 


the shaft. For purposes of illustration therefore we may rep- 
resent the excentric by a crank of equal throw. 

With this understanding let us again examine Fig. 209, on 
the left, when it is readily seen that the series of movements de- 
sired is exactly such as would be given by a small crank located at 
an angle somewhat more than 90 degrees ahead of the main 
crank. Or, in other words, with the valve stem attached to 
such a small crank or excentric, the valve will be so moved' that 
the piston will be forced to and fro in the manner described in 
Sec. 46 [i], and the main crank will follow the motion of the 
excentric. This is what is meant by saying that with a slide 
valve connected as in the diagrams, the excentric leads the crank. 

In regard to the angle between the crank and excentric it 
is clear from the diagram of Figs. 207, 209, that starting with 
the valve in mid position, it must move a distance equal to the 
lap before the port will begin to open. Hence when the piston 
is at the end of the stroke the valve must already have moved 
from its mid position by an amount equal to the lap plus the lead. 
Suppose now that the excentric is loose on the shaft and may be 
adjusted as desired. In Fig. 220 let C denote the crank on the 
center. Then suppose the excentric first located 90 ahead of O C 
as shown at O A, where O A=O B= excentric throw. Then neg- 
lecting the slight effect due to the obliquity of the excentric rod, 
the valve will be in its mid position, as shown in Fig. 207. Next, 
move the excentric ahead until the valve has moved a distance 
equal to the lap plus the lead as in Figs. 208, 220. This gives the 
final location of the excentric for the proper operation of the 
valve, as already described. The angle AOB through which it is 
thus necessary to move the excentric in order to affect this move- 
ment of the valve from its mid position, or more exactly the angle 
between the excentric and the line at right angles with the crank 
we shall term the angular advance. This term is sometimes used 
in reference to the entire angle between crank and excentric, but 
we shall prefer to understand by the term the angle as defined 
above. The angular advance is usually denoted by the letter tf. 
With the arrangement of gear shown in Fig. 220 the angle be- 
tween crank and excentric will therefore be 90+^ 

In regard to the direction of motion of the crank, it is clear 
that when the piston is at the end of the stroke and the crank on 
the center, the latter must start off in such direction as will in- 
crease rather than decrease the opening of the port for steam. 



IfJ the connections and arrangements of a valve gear are known 
no matter how complicated, this principle will always furnish an 
answer to the question as to the direction in which the engine will 
turn. Thus in the direct connected gear, as in Fig. 220, with the 
piston at the end of the stroke, the crank moves in the same di- 
rection as the excentric, or follows it, simply because that is the 
direction which opens the port still wider for steam admission. 

Fig. 220. Diagram Showing Connections for Simple Valve Gear. 

With an inside valve the excentric is set 180 behind or directly 
opposite the position for an outside valve, as shown in Fig. 221, 
or at OBj, Fig". 220. In such case the angle between the crank 
and excentric is 90 d or the angle $ is set off toward the crank 
from the 90 position. It is readily seen that this will bring the 
valve slightly open when the piston is at the end of the stroke, 
and that the crank will move leading the excentric because this is 

Fig. 221. Inside Valve and Location of Excentric. 

the direction which opens the port wider for steam admission. It 
is also seen that to fulfil the same purpose the piston and valve 
at the end of the stroke must move in opposite directions. 

In some cases the valve-rod instead of being directly con- 
nected to the excentric rod is worked through a rocker-arm, 
which reverses the motion as compared with the direct connected 
gear. See Fig. 222. This provides a second mode of variation 




Marine Engin.ef.nnj 

Fig. 222. Diagram for Valve Connections Through a Rocker Arm. 

which may affect the arrangement of a valve gear, giving in all 
four combinations, as shown in the following table : 

Valve. Connection. 



Angle Between Crank 
and Excentric. 
90 + A 
90 c 
90 (J 
90 -f 6 

Which Leads. 

Excentric Leads 
Crank Leads 
Crank Leads 
Excentric Leads 

[a] Oval Valve Diagram. 

We will now proceed to an examination of the effects due 
to varying the steam and exhaust laps, and the angle of advance 
[i]. This may be done most conveniently by the aid of a dia- 
gram. In Fig. 223 let AB represent to any convenient scale the 
path of the piston. Then from the various points of AB corre- 
sponding to a series of successive piston positions, let the dis- 
tance of the valve above or below its mid position be laid off, and 
the points thus found be joined by a continuous line. For a 
revolution or double stroke the result will be found somewhat as 
represented by the curved line of the diagram. Thus at the 
lower end of the stroke, say at B, the valve will be at a distance 
BT above mid position. When the piston has gone on to S it will 
be at a distance SJ,at R a distance RK, at P a distance PL, at 
N it will be in mid position, and will then pass below] and reach 
a distance AV at the end of the stroke A. On the return stroke 
the valve will pass through a similar series of locations deter- 
mined by the distances from AB downward to the curve. With 
this arrangement of diagram the movement of the valve above 
mid position is laid off above the line AB and vice versa. With 
an outside valve this shows above the line AB the events for the 
///> stroke and below the line the events for the down stroke. 



With an inside valve the relation is reversed, the events for the 
up stroke being shown below the line and for the down stroke 
above the line. 

We have already seen that starting from mid position the 
valve must move a distance equal to the steam lap before the 
port begins to open for steam. Suppose then, that BF is laid off 
equal to the steam lap on the lower end of the valve, and a line 
FL drawn parallel to BA. Then it is clear that the remaining 
distances from FL upward to the curve ZTKL give the dis- 

: /x s 

Fig. 223. Oval Valve Diagram. 

tances which the edge of the valve travels beyond the first edge 
of the port, and hence if the port is sufficiently wide, the actual 
widths of port opening. At R the farthest distance is reached, 
and the edge of the valve is at a distance OK beyond the first" 
edge of the port. It is clear that the distance RK is the throw of 
the excentric, and that this is equal to the lap plus the greatest 
width available for port opening. If the width of port is equal 
to or greater than OK, then the total movement OK can be 
utilized for opening -and will be its greatest value, while the vari- 


ous distances from FL to ZTKGL will give the entire history 
of the width of port opening for corresponding positions of the 
piston. Very often, however, the width of port is less than the 
distance OK. In such case draw a line IJ parallel to LF and at 
a distance from it equal to the width of the port. Then it is clear 
that the widths of port opening will be given by the distances 
from FL to the lines ZTJIL. Full opening is reached at J or 
with piston at S. This continues to I, or until the piston reaches 
Q. The port closes at L or when the piston reaches P. This is 
known as the point of cut off or steam closure. Similarly the 
port opens at Z just before the end of the stroke, and at the end 
is open a distance FT, which therefore represents the steam lead. 

Having thus obtained a general idea of the nature of this 
diagram let us examine by its aid the effects of a change in the 
steam lap. This corresponds to raising or lowering the line 
FL, and it is readily seen that the results of an increase, for ex- 
ample, are as follows : earlier cut-off, steam opening nearer the 
end of stroke, decreased lead, decreased port opening all the way 
through. The results of a decrease of lap are of course in the 
opposite direction respectively. 

In a similar manner we may examine the influence of a 
change in the exhaust lap. This is illustrated in the same figure 
where M Y is the lap line laid off at a distance B D from the 
center line, equal to the exhaust lap on the upper end of the 
valve; that is, on the end which opens the port to exhaust by 
moving above the mid position. It is thus seen that the valve 
opens to exhaust at Y with the piston a distance Y D from the 
end of the stroke, while at the end of the stroke it is open a dis- 
tance D T the exhaust lead. Then during the following stroke 
the distance of the exhaust edge of the valve from the corre- 
sponding edge of the port is given by the distance from D M to 
the curve T K L M. At M the port closes, and for M C, the 
remainder of the stroke, the steam undergoes compression in the 
cylinder. Usually the width of port is somewhat less than the 
total distance available, so that the full movement of the valve 
cannot be utilized for actual opening. In such case draw a line 
G H parallel to M Y and distant from it an amount equal to the 
width of port. Then full opening is reached at H with piston at 
W and continues to G with piston at U, the openings for the 
remaining portions of the stroke being given by the distance 
from M D to Y T H and G M. 



An increase of exhaust lap is thus seen to produce the fol- 
lowing results : earlier exhaust closure and longer compression, 
exhaust opening or release later or nearer the end of the stroke, 
decreased exhaust lead, decreased port openings or decreased 
time during which the port is wide open. A decrease of exhaust 
lap will of course, produce results in the opposite direction. 

We have thus far been concerned with the influence of an 
increase or decrease in the steam or exhaust lap. It remains to 
examine the influence due to a change in the angle 3 the angular 
advance. If this angle is increased and the new series of valve 
movements plotted as in Fig. 223, it will be found that the oval 
becomes narrower and touches the boundary lines nearer the 

Fig. 224. Oval Valve Diagrams. 

corners, as shown in Fig. 224. Similarly with a smaller value of 
d the valve oval will become wider or rounder, and will touch the 
boundary lines farther from the corners, likewise in the figure as 
shown. Remembering this and comparing Figs. 223 and 224 it 
is readily seen with the same values of the lap that changes in 
the various quantities will take place as shown in the tabular ar- 
rangement below. This table shows at a glance the variation in 
the various events due to change in the three items above the 
double line, as explained in the foregoing. 

3 7 6 


Attention may also be called at this point to the diagram of 
Pig. 225, in which the various quantities for an outside valve are 
lettered and named in position. A careful study of this figure in 
connection with the valve positions of Fig. 209 will be of great 
aid in acquiring a good understanding of the operation of a slide 
valve, and of its representation by means of such a diagram. 

If the measurements for the diagram are taken from an 
actual engine or from a properly constructed model, it will be 
found that for the down stroke the curve is more humped or 













































rounded than for the up stroke, as shown in the figures above. 
That is, for the same relative positions of piston the valve will 
be farther from the center line on the down than on the up stroke. 
This is an effect due to the angularity of the connecting rod of 
the engine, and it results that the various points of opening and 
closure, and the values of lap, lead and port opening cannot be 
made the same for both strokes. As is shown by the diagram 
the cut-off is usually later on the down than on the up stroke, 
and it cannot be equalized without a serious derangement of the 



other events. Instead of attempting to equalize any two feat- 
ures such as steam lead, cut-off or port-opening, it is usually bet- 
ter to so adjust the steam and exhaust laps above and below that 
the resulting combination of events shall represent the best com- 
promise possible under the circumstances. If it were not for the 
effect due to the angularity of the connecting rod, the curve 
would be an ellipse, and the distribution of the events for both 
strokes could be made the same. 






S O B=Steam Opening Bottom. 
E O T= Exhaust Opening Top. 
E C B=Exhaust Closure Bottom. 
S C T= Steam Closure Top. 

.'/urine Enginttring 

S C B= Steam Closure Bottom. 
E C T= Exhaust Closure Top. 
EO B= Exhaust Opening Bottom. 
S O T Steam Opening Top. 

Fig. 225. Oval Valve Diagram. 

[3] Bilgram Valve Diagram. 

In Fig. 226 let a circle be described with OP as radius equal 
to the throw of the excentric. Then draw the radius OP at an 
angle 6 (the angular advance) with the horizontal or line AB. 
Then draw CD above AB at a distance LM equal to the lead. 
Then from the point P as center and with PQ as radius describe a 



circle tangent to CD. Also on OP as a diameter describe a circle 
as shown. Then let A x be any position of the crank. Draw the 
radius AjO and extend it back to cut the circle on OP at E. 
Then the properties of this diagram are such that the movement 
of the valve from mid position is given by the distance PE, while 
the port opening will be less than this by PQ or PF the radius 
of the circle about P as center. This radius equals the steam 
lap, and the circle is for that reason known as the lap circle. It 
follows that EF is the port opening, or at least the travel of the 
edge of the valve beyond the edge of the port. The same con- 
struction holds for all other crank positions, so that we have here 
a means of representing by straight lines and circles the move- 

Fig. 226. Bilgram Valve Diagram. 

ment of the valve, and of thus determining the various events of 
the revolution. 

It must be understood, however, that while this construction 
shows with fair accuracy the relation between the movement of 
the valve and of the crank, it does not, without a further special 
construction, connect the movement of the valve with that of the 
piston. It is the special property of the oval diagram of [2] to 
show this latter relation for an actually constructed gear or for a 
model, where both sets of measurements may be actually made. 
It is the special property of the Bilgram diagram, and others 


composed of straight lines and circles, to connect together with- 
out actual measurement the movement of the valve and crank 
for the ideal case when there is no angularity of excentric-rod. 
The error here involved is usually small, and since the diagram 
is so easily constructed, it may be preferred for many purposes of 
initial design. 

It thus appears that the distances of the edge of the valve be- 
yond the edge of the port are given by the intercepts (shown by 
the shaded part of the diagram in Fig. 226), between the two cir- 
cles, one on OP as diameter and the other, the lap circle, about 
P as center. In case the width of the port is less than the dis- 
tance G O, an arc RS is drawn from P as center, such that the 
distance HS equals the width of port. The widths of opening 
are then given by the intercepts between J K S R L and the lap 
circle with P as center. It is seen that the port opens for steam 
at J with crank at A 2 and closes at L with crank at A 3 , while 
full opening holds from S to R. 

An entirely similar construction throughout with the ex- 
haust lap circle as shown by the small circle about P will give the 
various features of the movement on the exhaust side. 

The results of a change in the steam or exhaust lap, or in 
the angular advance d are readily examined by the aid of this 
diagram, and will be found to agree with the statements of the 
table in [2] . 

[4] etiner Valve Diagram. 

In Fig. 227 let ABCD be a circle described with radius OA 
equal to the throw of the excentric. Let A denote the angular 
position of the crank at the end of the stroke, and let OB be 
drawn perpendicular to AC. Draw OP at the angle <5 (the angu- 
lar advance) with OB, and on OP as diameter describe a circle as 
shown. Draw also an arc of a circle with OL equal to the steam 
lap, as radius. Let OA n be any position of the crank. Then 
the properties of this diagram are such that the travel of the 
valve from mid position is given by the distance OG, while the 
port opening will of course be given by EG the intercept between 
this circle and the lap circle MN. A similar construction holds 
for other locations of the crank, and it thus appears that steam 
opening will occur at N with the crank in the position ON 
while closure or cut-off will occur at M with the crank in the po- 
sition OM. By describing a circle with center at O and with 
radius equal to the exhaust lap, a similar construction gives the 


various features of the movement on the exhaust side of the 
valve. The history of the port opening for steam is thus given 
~by the intercepts between MN and the circle on OP, as shown by 
the shaded part of the diagram. 

In case the width of port is less than the distance LP an arc 
RQ is drawn from O as center such that the radial distance be- 
tween MN and RQ equals the width of port. The history of the 
port opening is then given in the same way as for the corre- 
sponding case in Fig. 226, as there explained. 

The results of a change in the steam or exhaust lap, or in 
the angular advance #, are readily examined by the aid of this 

Fig. 227. Zeuner Valve Diagram. 

diagram, and will be found in accord with the statement of the 
table in [2] . 


We have thus far examined in some detail the operation of 
a slide valve operated by a single excentric. In the course of 
the discussion it has appeared that with such a gear the direction 
of motion of the crank relative to the excentric depends on 
whether the valve takes steam on the inside or outside, and on 
the nature of the connection between the excentric rod and 
valve stem. With any one arrangement, however, motion in one 
direction only is possible ; and it is therefore clear that to enable 


the engine to reverse a fundamental requirement of all marine 
valve gears some additional features will be necessary. 

The general problem of a reversing valve gear is one which 
has been solved in a great variety of ways, both with and with- 
out the use of excentrics, as we shall see later. With the use 
of excentrics the simplest solution is furnished by the well-known 
Stephenson link. This is illustrated geometrically in Fig. 228. 
C represents the crank on the center. A denotes one excentric 
at an angle COA on one side of the crank and B another at the 
same or approximately the same angle COB on the other side. 
AD and BE denote two excentric rods connected by a link DE 
curved to an arc whose radius is the length of the rod AD or BE. 
To a block in this link is attached the valve stem DG. For the 
structural details of this -gear reference may be made to Sec. 53. 

Marine Engineering 

Fig. 228. Skeleton of Stephenson Link. 

Now it will be readily seen that with the arrangement of the 
diagram the valve is under the control of the excentric A alone. 
The excentric B simply pulls the link DE back and forth, causing 
it to swing about D, but in no way affecting the movement of 
the valve. Hence the engine will move entirely under the con- 
trol of the excentric A, and with an outside valve direct con- 
nected, the direction of rotation will be right-handed, or from C 
toward A. Now to effect the reverse, it is only necessary to 
move the link over so that E comes to the center line and the 
valve and engine pass under the control of the excentric B alone. 
For the reasons already noted in Sec. 47 [i] the motion will 
now be reversed and the rotation will be left-handed or from 
C toward B. 


We have next to tequire regarding the motion of the valve 
stem when the link is only part way over, as shown by the 
broken lines in the diagram. In such case it is readily seen that 
the motion of the valve will be derived partly from the excentric 
A and partly from B, the former giving the principal part ot 
the motion, and the latter exercising a modifying influence. Now 
without taking up the examination of this question in detail, it 
will be sufficient to state that within a very small error the result- 
ant motion will be the same as though it were given by a single 
excentric of somewhat decreased throw ,and increased angular 
advance as compared with A or B. A simple construction will 
serve to determine the throw and angular advance of this equiva- 
lent single excentric. This may be carried out as follows : 

(1) In Fig. 229 lay off the two excentric throws OA and OB, 
with the proper angular advance as shown, and draw the line AB. 

(2) Divide the length of the link DE Fig. 228 by twice the 
excentric rod AD and multiply the quotient by the length AC 
Fig r ' 229. 

(3) Lay off the result from C to D, and through the three 
points A D B pass the arc of a circle, as shown. 

Then the arc ADB may be considered as representing the 
link, and to find the throw and angular advance of the equivalent 
excentric for any given position of the link-block as F on D l E l . 
Fig. 228, we have only to take a corresponding point F on the 
arc ADB, and draw the radius OF. The throw is then repre- 
sented by OF and the angular advance fi by the angle POF. 

That is, if the link be put into the position shown by broken 
lines in Fig. 228 the movement of the valve will be, within a small 
error, the same as though it were operated by a single excentric 
of throw OF, Fig. 229 and angular advance POF determined 
in the manner described. It is readily seen, therefore, that as 
the link is so moved as to bring the block from the end nearer 
and nearer the center, the corresponding point F Fig. 229 will 
move from A nearer and nearer to D, and the throw of the 
equivalent excentric will continually decrease while the angular 
advance will increase. As the link passes the center and the 
block approaches the other end the corresponding point F moves 
on from D toward B, and the throw again increases, while the 
angular advance changes to the other side and gradually de- 
creases as B is approached. With the link in full gear at either 
end, the corresponding point F comes to either A or B, and the 



equivalent excentric becomes the same as the real excentric, 
with its throw and angular advance as constructed. 

In some cases, especally with the double-bar form (see Sec. 
53) the link may be put over so as to bring the block even be- 
yond the points of attachment of the excentric rods. See Fig. 
243. In such case the throw and angular advance of an equi- 
valent excentric will be given by extending the arc beyond 
A and B as shown in Fig. 229, and by then taking a point F cor- 
responding to the relative location of the block and link. In such 
case it is seen that the throw is increased and the angular ad- 
vance decreased. 

The details of the motion for any given position of the link- 
block may, of course, be determined by the use of any of the 

Fig. 229. Construction for Equivalent Excentric Stephenson Link. 

methods given above for the case of a single excentric. It is 
simply necessary to take the equivalent throw and angular ad- 
vance determined as in Fig. 229, and use them according to the 
methods described in Sec. 47. In this way it may be found that 
as the gear is linked up, or the block approaches the middle, the 
valve^ travel and port openings decrease, the lead increases, and 
the cut-off becomes earlier. This is further illustrated by the two 
oval diagrams" of Fig. 230. These are constructed as explained 
in Sec. 47 [2], and represent the effect of linking up. The 
larger diagram represents the movement for full gear position 
as shown by the full lines of Fig. 228, while the smaller and 
narrower one represents that for a linked up position as shown 
by the broken lines of the same diagram. 



A gear arranged as in Fig. 228 is known as an open gear or 
gear with open rods. That is when the crank is turned away 
from the cylinder the rods are open as shown. If instead of 
this the rods are crossed as shown in Fig. 231, then it is called a 
gear with crossed rods. It will be noted that with the open gear 
the rods become crossed when the crank is turned toward the 
cylinder, while in the same position the rods in the crossed gear 
become open. It is therefore necessary to note the character of 
the gear by the appearance of the rods when the crank is 
turned away from the cylinder as stated above. It must now 
be remembered that the construction given in Fig. 229 and the 

Fig. 230. Oval Valve Diagrams for Stephenson Link. 

conclusions drawn from it apply to the gear with open rods 
only. For the crossed rod gear, however, a similar construc- 
tion applies as shown in Fig. 232. The distance CD is here laid 
off toward the center and a like arc is passed through the three 
points ADB as shown. The diagram thus constructed is used in 
the same manner as with Fig. 229. It is thus seen as the link- 
block approaches the center of the link and the corresponding 
point F approaches D, that the equivalent angular advance in- 
creases, the equivalent throw, valve travel and port openings 
decrease, even more rapidly than with the open gear, while 
the lead decreases and the cut-off is earlier and earlier. 

The principal characteristics of these two types of Stephen- 
son link valve gear with regard to the effect on the various 
events, etc., due to linking the gear up may be conveniently pre- 
sented in the following tabular form : 





Type of Gear. 

Open Rods. 

Crossed Rods. 

Equivalent Excentric Throw Decreased 

Angular Advance Increased 

Valve Travel Decreased 

Port Opening Decreased 

Lead Increased 

Cut-Off Earlier 


Fig. 231. Skeleton of Stephenson Link, Crossed Rods. 

Fig. 232. Construction for Equivalent Excentric with gear of Fig. 231. 

The chief point of difference is seen to be in the lead, which 
increases with open rods and decreases with crossed rods as the 
gear is linked up. While both types of gear are met with, the 
open rod gear is more frequently employed. This is due to the 
fact that as the cut-off is made earlier by linking up, an increase 


of lead may be preferred to a decrease, and also to the fact that 
the width of port opening is decreased less rapidly by the open 
than by the crossed gear. 


The Stephenson link is by no means the only form of valve 
gear which will allow of reversal and which will give variable 
cut-off. Among the other arrangements are several known as 
"radial valve gears," and of these the more important may be 
briefly described. 

In Fig. 233 let C denote the position of the crank and E 
the position of a single excentric located directly opposite the 
crank. Let FD be a slide pivoted on the horizontal line OX, 
and EB a rod attached to the excentric at E and fitted with a 
pin joint and block at P so that the block may move in or A on 
the slide FD. Then as the excentric moves around O the end 
E of the rod describes a circle, the point P describes a straight 
line FD back and forth, and other points between P and E 
describe paths intermediate between these two. For a point 
such as Q this is found to be an inclined oval as shown in the 
diagram. For points beyond P such as Q', for example, the 
path is found to be a somewhat similar oval as shown. Now 
it is found that the proper motion for a valve can be derived 
from a point moving as Q or Q' and that it is simply necessary 
to connect such point by a proper link to the valve stem as 
shown for Q. 

Instead of placing the excentric at 180 degrees from the 
crank it may be placed with the crank, in which case in Fig. 233 
we should consider the crank at C'. There are thus four arrange- 
ments of the gear according as the excentric is with or opposite 
the crank, and as the point Q is between P and E or beyond P. 

In all cases the gear must be so adjusted that when the 
crank is on either center as C, the point P is on OX, or at the 
pivotal point of FD. It is readily seen that if this condition is 
fulfilled for one center it will be likewise for the other. 

With regard to the kind of valve to be used (inside or out- 
side) and the direction in which the engine will run for any given 
arrangement of gear, the same principles may be applied as in 
the case of the single excentric. Thus it is easily seen from the 
symmetry of the motion that Q will go as far above the center 
line as below, and hence that OX contains the middle of its 



vertical motion. Hence in the arrangement of the figure with 
the crank on the top center the valve is below its mid position 
by the distance from Q to the center line OX. Hence to take 
steam on the top of the piston an outside valve must be em- 
ployed, and the engine will go in the direction which will open 

Fig. 233. Braemme Marshall Radial Valve Gear. 

the valve still wider. This must be that which ^will lower the valve, 
and hence that which will lower P, and hence that which will 
carry E to the left, or right-hand rotation. If, on the other 
hand the motion were derived from Q', the valve will be above 


its mid position, and hence to take steam on top it must be an 
inside valve. The engine in this case will turn in the direction 
which will raise Q' still further. This requires E to move to the 
right, and hence the rotation will be left-handed. 

If FD be inclined in the other direction, as shown by the 
dotted line, it will be readily seen by the same rule that the 
direction of rotation in each case will be reversed. It is also 
found that as the direction of FD approaches the horizontal or 
OX, the cut-off becomes earlier and earlier, and it is readily 
seen that the valve travel and hence the port opening will de- 
crease. We have here therefore an entirely similar action to 
that which takes place in linking up with the Stephenson link. 
The means for changing the cut-off and for reversal are there- 
fore furnished by providing a means for changing or reversing 
the obliquity of the slide FD, and for retaining it in any position 

Since the point P comes to the center of the slide or pivot 
point when the engine is on the center, it follows that in this 
position the line EB, the point Q and the valve will have exactly 
the same location no matter what the position of FD, and hence 
no matter what the point of cut-off. Hence the lead of the 
valve will be the same for all points of cut-off, or in other words, 
the lead is not affected by the change in cut-off. This is a 
feature which, as we shall see, is possessed by the various forms 
of radial valve gear. With the Stephenson link the lead is vari- 
able with the cut-off as described in Sec. 48. 

As noted above there are four arrangements of this gear 
depending on the location of the point Q, and the angle between 
the excentric and the crank. 

Location of Q. Valve. 


Inside P 



Inside P 



Outside P 


1 80 

Outside P 


These four arrangements are shown in the above table 
with the appropriate form of valve. The direction of rotation 
in each case will depend of course upon the direction of obli- 
quity of the slide FD. 



The arrangement of Fig. 233 shows the earlier form of the 
gear. Another and later form is shown in Fig. 234 in which 
the slide FD Fig. 233 is replaced by an arm Pl\ pivoted at T l9 
and attached by a wrist pin to the rod EB. In this way the point 
P is caused to move in the arc of a circle F l D l inclined to the 

Fig. 234. Braemme Marshall Radial Valve Gear. 

horizontal OX as shown. In such case the motion of the valve 
is very nearly the same as if the point P moved in the line of the 
tangent F D, and except for secondary modifications it is there- 
fore equivalent to the motion of Fig. 233. The change of cut-off 
and the reversal are brought about by swinging the pivot T x 



about a center S, the intersection of FD with OX. In this way 
Tj is brought to a new position T 2 and the line of motion is 
brought to F 2 D 2 , thus reversing the direction of rotation ac- 
cording to the same principles as applied in the preceding case. 




N \ i 

1 \ 


i 1 


1 1 



Fig. 235. Joy Radial Valve Gear. 

We may evidently have here the same four arrangements as in 
the other form of the gear, with the same relation between ex- 
centric angles and type of valve as given in the table above. 


Sec. 50. JOY VAI,VE GEAR. 

In this form of radial valve gear no excentric whatever is 
required. As shown in Fig. 235 F is a point on the connecting 
rod to which a link DF is attached pivoted to a swinging or sus- 
pension bar DK. The point F moves in an oval path as shown. 
The point D moves in the arc of a circle as shown. An inter- 
mediate point as E will move in a path somewhat as shown. 
From this point on, the gear is similar 1 to a Braemme-Marshall. 
Thus EP is a link pivoted at E and with the point P carried by 
a suspension bar PR exactly the same as PT of Fig. 234. The 
motion for the valve is similarly taken from a point Q as shown. 
It is thus clear that the Joy gear as here shown is the gear of 
Fig. 234 in which, however, the point E instead of moving in a 
circular path derives its motion from the connections shown in 
Fig- 235 and moves in a distorted oval path as there shown. 
It is clear that there will be the same two varieties of gear ac- 
cording as the point Q is taken between E and P or beyond P. 
The reversal is also effected in the same fashion by swinging 
PR, or the straight slide may be used as in the gear shown in 
Fig- 233. 


In this gear as shown in outline in Fig. 236, one excentric 
E is used. MGK is a curved link pivoted at G. H is a block 
sliding on or in the link and connected by a radius arm 
to the valve lever AF. The end F of this lever is connected by 
a pin joint to the valve rod, and the other end A by a short link 
to the main crosshead as shown. The valve is thus seen to de- 
rive its motion in part from the main crosshead and in part from 
the excentric. The former part comes through the valve lever 
which pivots about D and thus communicates motion from C to 
F. The latter part comes from the curved link which is operated 
by the excentric, causing it to swing about G and thus through 
the radius arm and valve lever the motion of the excentric is 
communicated to the valve rod. The combination of these two 
motions is found to be such as to give a suitable movement to 
the valve. 

The linking up and reversal are accomplished by swinging 
the block H and radius rod from one side to the other of the 
mid-position or pivot G. As the block H is brought nearer to G 
the cut-off becomo* earlier while the valve travel and port-open- 



ing are decreased the same as with the Stephenson link above 

In order that this gear may operate properly, certain ad- 
justments are required as follows: 

The radius of the curved link MGK must equal the radius- 
rod HD. 

Place the crank on say the top center, and bring the link 

Fig. 236. Walschaert Valve Gear. 

MGK into line with an arc struck from D as center with DH 
as radius. Then M should be so taken that the angle OMG is 
a right angle. Also the excentric E is placed at a right angle 
from OM as shown. Then in this position the block H may 
be drawn to and fro along the link without moving the valve. 
It is also clear that after the crank has gone 180 degrees the 
excentric will be at E, and MG will be again in the same position 


and the block H may be drawn along the link without giving 
motion to the valve. It follows that the position of the valve 
when the crank is on the centers will be the same no matter 
where the block H may be located, and hence that the lead 
of the valve will be the same for all points of cut-off. 

It is clear that the arrangement of Figs. 233-236 will bring 
the valve chest on the side of the cylinder transversely instead 
of fore and aft. The cylinders may therefore, so far as valve 
chests are concerned, be placed nearer together than with the 
Stephenson link, in which case the valve chests are forward or 
aft of the cylinders. The length of the engine as a whole may 
therefore be made somewhat less with the radial types of gear 
than with the Stephenson link, and it is this feature which gives 
to them their best claim of advantage. Such gears possess 
also, as we have seen, the property of giving the same lead for 
all points of cut-off, while with the Stephenson link the lead 
varies with the cut-off. With proper design, however, the varia- 
tion in the latter case is not sufficient to constitute a feature of 
any importance, and the difference on this point can hardly be 
considered as forming any noteworthy advantage for the radial 
gears. The general character of the valve movement and the 
distribution of events in all cases is best studied by the aid of a 
diagram such as that of Fig. 223. It will thus be found that the 
results for these various cases will be quite similar, and that 
such differences as appear are of relatively small importance. 
Speaking broadly, it is perhaps fair to say that there is not suffi- 
cient difference in the operation of the valve itself to furnish 
any pronounced claim of advantage for the usual cases arising 
in marine practice. The choice must therefore be made rather 
by reason of structural considerations, such as the shortening 
up of the engine referred to above, or the details of construction 
of the gear as affecting the questions of breakage, wear and 
tear, readiness of repair, readjustment, etc. On the whole, the 
Stephenson link seems to be usually preferred as the best ful- 
filling the all around requirements for the marine valve gear, 
and it may be fairly considered as the representative gear in 
present-day marine practice. 


As we have seen in Sec. 47 [i] the action of an excentric 
is equivalent to that of a simple crank of throw equal to the 



excentric and set at a corresponding angle with the main crank. 
It is evident then that a series of cranks will operate a valve 
and gear in a manner identical with a corresponding series of 
excentrics. Such cranks, however, on account of their small 
throw cannot readily be located or formed on the main crank- 
shaft, and hence where used for operating the valves, are neces- 
sarily placed on a special or auxiliary shaft. In Fig. 237 is shown 
the usual way of arranging the parts of this gear. 

S is the center of the main crank-shaft, and ST the main 
crank. A, B, and C are gear wheels, A attached to the crank- 
shaft, B an idler, and C attached to the valve shaft. O is the 
center of this shaft and OP represents the throw and angulai 
location of the small crank for operating the valve. The valve 

Fig. 237. Crank Valve Gear. 

shaft will then turn in the same direction as the main shaft, and 
as may be readily seen, will operate the valve precisely in the 
same manner as an ordinary excentric of the same throw and 
angular location. For reversing with this gear the usual plan 
is to have but one crank and valve-connecting-rod correspond- 
ing to one excentric rod. The angular location of the crank 
must then be changed from a position such as OP in the figure 
to OPj. By comparing this with Sec. 47 [i] it is seen that such 
a change in the location of the crank will necessarily cause the 
engine to move in the opposite direction. To bring about this 
change in the location of the valve crank relative to the main 
crank, various mechanical devices may be used. 

Thus it may be seen that if after the engine has stopped 



the gear C were slipped out of the mesh with B, turned around 
through the angle POP l and then slipped back into mesh again, 
the crank would be brought to OP l and if the engine were 
started again it would go in the opposite direction. This, of 
course, is not a practicable form of reverse, since it cannot be 
carried out quickly enough, nor when the engine is in motion. 
It does, however, serve to illustrate the necessary change to be 
made in the angular location of OP. 

In the usual mode of operation, some form of spiral cam is 
employed, as illustrated in Fig. 238. C is the gear wheel car- 
ried on a sleeve AB and connected to it by a key way and 
feather so that the sleeve may be moved back and forth axially 
and still remain coupled to the gear C so far as rotary motion is 
concerned. The gear C is also prevented by suitable stops from 
being carried out of mesh with the gear B, Fig. 237. This sleeve 
is carried on the shaft D to which is attached the valve crank, 

Fig. 238. Crank Valve Gear, Reversing Arrangement. 

and is loose on D and only connected with it by a pin P which 
projects through a spiral groove RS cut in the metal of the 
sleeve. At E there is a circumferential groove in the sleeve, in 
which is fitted the end of a controlling lever F, by means of 
which the sleeve may be moved back and forth longitudinally. 
In the figure the sleeve is shown pushed in so that the pin is at 
one extremity of its travel in the spiral groove, and the valve- 
crank will therefore be in a corresponding position with refer- 
ence to the gear C and main crank, which we will suppose to be 
for full gear ahead. If now the sleeve be pulled longitudinally 
until the other end of the groove contains the pin, it is clear 
that we shall have changed the angular location of the pin rela- 
tive to the gear and hence of the valve-crank relative to the 
main shaft. If then the spiral groove be of suitable extent and 
location, such a change will serve to move the crank OP, Fig. 


237, from its position for full gear ahead, to OP^ its position for 
full gear astern. Instead of one pin and spiral groove, two on 
opposite sides may be used, and other details may vary in many 
ways, but the arrangement will serve to illustrate the principles 

While this form of valve gear is thus efficient for reversing, it 
is much less suitable for linking up or varying the cut-off than 
the other forms of gear discussed above. Referring to Sec. 48 it 
was there shown how to find the equivalent simple excentric for 
any adjustment of the Stephenson link, as shown by the line 
A D B, Fig. 229, for an open rod gear. Now it is readily seen 
that the form of reverse just considered is equivalent to taking 
the excentric O A and carrying it around from OA to OB, so 
that for the varying intermediate positions the virtual excentric 
would be. given by drawing a line from O to the arc A B with 
OA as radius. From the principles discussed in Sec. 47 [2] it 
is readily seen that the effect on the valve will be as follows : 

Linking up will produce : 

Earlier cut-off. 

Earlier steam opening. 

Greatly increased lead. 

Earlier exhaust opening. 

Earlier and greater compression. 

The same valve travel and opening. 

While therefore the port opening is not decreased by link- 
ing up, the lead and compression may become excessive, and re- 
strict the practicable range of variable cut-off to narrow limits. 

In Fig. 237 we have referred to one cylinder and valve op- 
erating crank only, but the same arrangement may be applied, of 
course, to a series of cylinders for a multiple expansion engine. 
In such case the valve operating shaft has a series of cranks, one 
for each cylinder, and set at the proper angle from the corre- 
sponding main crank. Then with a reverse gear similar to that 
described in Fig. 238 all the cranks are moved together and are 
brought into the proper relation with their main cranks to oper- 
ate the engine in the reverse direction. 

It is also to be noted that this arrangement brings the valve 
chests at the sides of the cylinders and that on this account it has 
the same effect in shortening up the engine as the use of a 
radial valve gear. This type of gear has been used to some 
extent on launch, yacht and torpedo boat engines, but has not 



found favor for larger or slower running engines as used in ordi- 
nary mercantile and naval practice. 


In the present section we shall give a brief description of 
the more important details of the Stephenson Link valve gear as 
a representative gear, and including as it does most of the ele- 
ments of other forms of gear as described in Sec. 49-52. 

fi] Excentric and Strap and Excentric Rod. 

The general construction of this part of the gear is shown 
in Figs. 239-242. The excentric consists of a disc or sheave of 




Excentric, Detail of Construction and Fitting. 

circular form, and placed on the shaft excentric or with its cen- 
ter set out from the center of the shaft. As shown in the figure 
the sheave is made in two unequal halves which join on the cen- 
ter line of the shaft, and are secured by bolts as shown. This 
brings the center of the disc at the point shown, the distance AB 
between this and the center of the shaft being the excentric 
throw as defined in Sec. 47. 'Once adjusted the excentric is 
keyed in place and in addition set screws or binding bolts may 
be fitted, as shown. In the excentric of Fig. 239 the larger part 
of the sheave is lightened out to save weight. In excentrics of 
moderate size this is not usually done, and the bolts connecting 

39 8 


the two parts are tapped into the larger part and hold the other 
portion in place by means of a countersunk head. The material 
of the excentric is either cast-iron or steel, or if small in size, 
^brass is sometimes employed. 

The strap which surrounds the excentric as shown in Fig. 
240 is also made in two parts bolted together, and the excentric 
rod is attached by a flanged foot to the upper half as shown in 
the figure. There are several ways of fitting the surfaces of the 
strap and excentric together, as shown in Fig. 241. In modern 

Fig. 240. Excentric Strap, Detail of Construction and Fitting. 

practice the strap is usually of cast-steel or brass, lined with 
white metal for a bearing surface as shown at a and described in 
Sec. 21 [n]. 

With this arrangement of excentric sheave and strap, it is 
readily seen that as the former turns with the shaft the center 
of the sheave turns about the center of the shaft ; also that the 
center line of the excentric rod which will always pass through 
the center of the sheave, will therefore move exactly as though 
it were a connecting rod with the distance A B between the two 



centers as a crank arm. Hence as noted in Sec. 47 [i] the mo- 
tion communicated to the rod will be exactly the same as would 
be given by a crank and connecting rod the former of throw equal 
to that of the excentric. The other excentric strap and rod are 

Fig. 241.' Different Methods of Fitting Excentric Strap. 

fitted up in the same manner and the two rods connect with the 
ends of the link in the manner shown in Fig. 242. In this figure is 
shown on the left the lower end of the excentric rod with 
flange for securing to the upper strap as in Fig. 240. On the 

Fig. 242. Excentric Rods, Fittings at Ends. 

right is shown the upper end of the rod sometimes known as 
the excentric rod fork. The form is more properly that of a U, 
the two sides being fitted with bearing brasses and cap to pro- 
vide a bearing for the pins by means of which the connection 
is made with the link. 




In Fig. 243 is shown the usual form of double bar link as ft 
is termed. It consists of a pair of bars curved in the arc of a 
circle of radius equal to the geometrical length of excentric rod ; 
that is, equal to the distance from the center line of the link to 
the center of the excentric sheave. These bars are connected at 
the ends by bolts and by a block between so as to main- 
tain the desired distance between them. Near the ends are fitted 
the pins A, B, C, D, which serve for connecting with the excen- 








o O 








; ! 


O fo)) O 





Fig. 243. Stephenson Double Bar Link. 

trie rods by means of the bearings in the upper ends, as shown 
in Fig. 242. Another pair of pins E, F, is fitted, either at the 
center as shown or as an extension of the pairs at the ends, or at 
some intermediate point. To these are attached a pair of bars or 
links usually known as side or bridle rods as shown in Fig. 244. 
These lead to the rock or weigh shaft, and serve to control the 
gear when linking up or reversing, and to hold it in any desired 

The rock-shaft or weigh-shaft is usually carried in bearings 
on the outside of the columns near the top, and is provided with 



arms, one each for the several links, and one for the connection 
to the reverse cylinder by means of which it is operated as de- 
sired. This is sufficiently illustrated in Figs. 100, 116, 247. 

In order to provide an independent adjustment for the valve 

Marine Eiiyineering 

Fig. 244. Side or Bridle Rods. 

gear of the different cylinders, the weigh shaft arm as shown in 
Fig. 245 may be provided with a slot within which moves a 
block under the control of a hand-wheel and screw as shown. 
The bridle rods are attached to pins on the sides of this block 

Fig. 245. Independent Adjustment for Cut Off with Stephenson Link. 

as shown on the right, and by this means without moving the 
weigh shaft at all, the links may be given an adjustment within 
the limits of the motion of the block in the slot. It is custom- 
ary to so adjust the line of motion of this block that when the 



g:ear is in the ex>-ahead position, it shall lie nearly in the line of 
the bridle rods so that any movement of the block will be com- 
municated to the link without loss. In the backing position 
on the other hand, the line of movement of the block will lie 
across the line of the bridle rods at a considerable angle, and 
movement of the block back and forth will give but slight motion 
to the link. See also Figs. 100, 247. 

This arrangement is often of use in adjusting the points of 
cut-off in the separate cylinders so as to divide as equally as pos- 
sible the power among the different cylinders. See also Sec. =55 


[3] I/ink Block and Valve Stem. 

The connection between the link and the valve stem is 
made by means of a link-block as shown in Figs. 243, 246. This 






S .--<.. cP 

1 ' 

,-.-t~. \ i 



qS ] ; - '3s 


\ : \ 

! J! 


T ' T I 

i Jfari'K Enyineering\ 

Fig. 246. Link Block for Stephenson Double Bar Link. 

consists of a central pin with wing pieces at the ends, as shown 
in the figure, the latter being fitted with bearing surfaces for 
connecting with the bars of the link, and permitting sliding mo- 
tion between the two. 

The pin is connected with the lower end of the valve stem, 
which is formed in the usual manner with brasses and cap. See 
also Fig. 248. The valve stem is usually guided by means of a 
special guide or bearing as shown in Figs. 100, 247, which sup- 
ports it against side stress, especially at the stuffing-box just 
above. In good practice the valve-rod stuffing-box is usually 
packed with some form of metallic packing, and is of the same 



general form and arrangement as the piston-rod stuffing-box 
described in Sec. 24 [7]. Passing through the stuffing-box the 
valve-stem is attached to the valve, and thus the chain of con- 
nections between the excentric and the valve is completed. 

The assemblage of these various parts of a Stephenson 
valve gear is further illustrated in Fig. 247, showing the upper 
ends of the excentric rods, the link, link block, valve-sttm and 
guide, bridle rods, rock-shaft arm and brackets for supporting 
the shaft, and independent cut off control in rock-shaft arm. 

Marine Engineering 

Fig. 247. Arrangement of Stephenson Link and Rock Shaft Connections. 

Reference should also be made to Figs. 97, 99, 100 for further 
general illustrations of this gear. 

When two piston valves are driven side by side as is very 
commonly the case on the L. P. or I. P. cylinders, the two valve- 
stems are connected across by a yoke as shown in Fig. 248, 
which in turn is connected to the link block by a form of bearing 
similar to that for the single stem. In such case the guide is 
very commonly attached tp the yoke, the arrangement consist- 



ing of a dovetailed or gibbed slide and guide, the first formed on 
the yoke, and the second by a vertical plate or bar projecting 


Vi # 1 2 3 4 5 

Fig. 248. Yoke and Guide for Driving Double Piston Valves. 

downward from the bottom of the cylinder head as shown in 
the figure. 




[i] Putting an Engine on the Center. 

One of the important features of valve setting is the placing 
of the engine on the centers or dead-points in order to deter- 
mine the lead. In a rough way this may be done by turning the 
engine and watching the cross-head slide as it approaches the 
dead-points. The slide will move along the guide, more and more 
slowly, and will finally stop and begin to return. Just as the far- 
thest point is reached, the crank is on the dead-point. By mov- 
ing the engine back and forth and watching carefully the move- 
ment of the slide relative to a light mark or score on the guide, 
the desired point may be determined with fair accuracy for pur- 
poses of valve setting. 

The difficulty in making an accurate determination by this 
method lies in the fact that when near the center the crank mav 

Marine Engi 

Fig. 249. Putting an Engine on the Center. 

be moved to and fro through a sensible angle with hardly a no- 
ticeable movement of the slide. Hence while it is possible to 
determine to a nicety the point on the slide which corresponds 
to the highest or lowest position of the piston, it is less easy 
to know just where to set the crank so as to have it accurately 
correspond to the same location. For more accurate setting the 
following method may be used. 

The engine is placed with the cross-head slide at a sma'll 
distance from the lowest or highest position. A mark A is then 
made on the slide and "a corresponding mark B on the guide. 
The distance which the cross-head should be placed from the 
center when these marks are made depends of course on the 
size of the engine, but I or 2 inches or say 1-20 to i-io the stroke 


will be a suitable distance. See Fig. 249. Another pair of 
marks P, Q, is next made on the forward end of the shaft and 
the adjacent brass, or on one of the coupling flanges and an ad- 
jacent block or bar set up for the purpose. The object is sim- 
ply to have two pairs of marks, one on the cross-head slide and 
its guide, and one on the shaft and its support or guide or an ad- 
jacent and fixed object. The engine is then moved around con- 
tinuously in one direction past the center till the cross-head slide 
moves back and the mark A again comes opposite B. The 
point P on the shaft in the meantime will have moved on to a 
new location, and a corresponding mark R is made on the bear- 
ing or stationary part of the engine on which the first mark Q 
was placed. The angle between these two marks Q R, on the 
shaft or coupling corresponds to the movement of the cross-head 
from the position of the first pair on to the end of the stroke and 
back again an equal distance. A mark S midway between Q 
and R, will give the proper location for the mark P when the 
cross-head is at the end of the stroke and the crank on its dead 
point. In this way by moving the shaft till P is brought oppo- 
site S the location of the crank for each dead point may be quite 
accurately found. 

[2] Setting the Valve. 

To return to the setting of the valve we first note that the 
distribution of the steam to a cylinder by means of a slide valve 
depends on four chief items : 

(1) The throw of the excentric. 

(2) The angular location of the excentric relative to the 

(3) The length of the valve stem. 

(4) The steam and exhaust laps. 

Now let us assume that the parts of the valve gear are made, 
and that it simply remains to connect them up and make the 
proper adjustments. It is seen that we have but two items 
which may be varied, or which may enter into the question of the 
setting of the valve. These are (2) and (3) above. We should 
first adjust for (3) and then for (2). 

Referring to Sec. 55 [2] (9) it appears that an incorrect 
length of valve-rod will give an improper balancing up of the 
events for the top and bottom of the cylinder. Hence the vari- 


ous events for both ends must be examined and compared. To 
this end the entire gear is connected up according to judgment, 
the link being placed in the position intended for normal running 
ahead, and the necessary arrangements made for observing the 
movement of the valve. If the valve is a flat slide the valve 
chest cover is left off for this purpose. With piston valves, 
however, it is necessary to observe the movement of the valve by 
means of peep holes through the shell of the chest, such holes 
being fitted with screw plugs or covered by caps when the en- 
gine is closed up and ready for operation. In this manner the 
lead is observed while the engine is on fHe centers, and the 
points of cut-off and other items are observed for each end of the 
cylinder. The location of the valve on the stem is then varied 
until a fair balance between the two ends is obtained. It will be 
found that with anything like equal leads the cut-off will be later 
on the down than on the up stroke, or with an attempt to even 
the points of cut-off the lead and port opening on top will be- 
come too small and the lead on the bottom excessive. It will 
often be found that with something approaching equal leads on 
top and bottom, the points of cut-off will vary in the two ends 
by nearly or quite 10 per cent, or even more. 

It is readily seen that similar derangements will result from 
an attempt to balance the exhaust items. In general it is far 
better not to attempt to exactly balance any one item in the two 
ends, but simply to aim for the best all around combination of 
events which can be obtained in the given case. 

When a fair balance is thus obtained, the question of the point 
of average cut-off, steam-opening, release and compression may 
be taken up. Changes in these items require a change in the 
angular location of the excentric relative to the shaft, and it may 
be shifted according to the relations shown in the table of Sec. 
47 [2] until the general character of the various items is made 

It thus appears that of the two items, length of valve-rod and 
location of excentric, the latter really fixes the general character 
of the various items, while the former makes it possible to ap- 
proximately even up or balance the various items between the 
two ends, according to what seems the most desirable average 

If the excentric is shifted through any considerable an.rle 
from its first location it will be necessary to again examine the 


question of balance between the two ends, and to again adjust 
the length of valve-rod. 

If by the adjustment of both valve-rod and excentric, the 
desired events, openings, etc., cannot be obtained, it means that 
the trouble lies with one or both of the other items, throw of ex- 
centric and steam or exhaust lap, and steps must be taken to 
modify these features as the conditions may require. 

The following table gives an illustration of the balancing up 
of the various items in the two ends of the cylinder. 


Steam opening 

.7 per cent, before 
end of stroke 

.4 per cent, before 
end of stroke 

Steam closure or cut-off 

68 per cent. 

59 per cent. 

Exhaust opening 

90 per cent. 

91 per cent. 

Exhaust closure 

85 per cent. 

84 per cent. 

Steam lap 

2.44 in. 

2.40 in. 

Exhaust lap 

.12 in. 

+ .68 in. 

Steam lead 

.60 in. 

.52 in. 

Port opening for steam 

2.08 in. 

2.12 in. 

Angle of advance 

33 degrees 

Throw of excentric> 

4 15-16 in. 

The setting of the valve may of course be examined or re- 
adjusted at any time, as desired, by the use of the same general 

[3] Valve Setting from the Indicator Card. 

The indicator cards interpreted in accordance with the re- 
lations given below in Sec. 55 [2] furnish most valuable evidence 
as to the adjustment of the valve gear, and its suitability for 
operation under steam. In attempting a readjustment or re- 
setting by the aid of the indications given by the cards, the 
question of balance between the two ends as affected by the 
length of valve-rod should be taken first, next the items depend- 
ing on the angular location of the excentric, and last the ques- 
tion of lap and excentric throw. 

If the cards are pushed over to one side or show differences 
in the two ends as in Fig. 260 it is evidence that the valve stem is 
not of the right length, and it must be changed accordingly. 


This is done first, cards being taken after each change until 
the two ends are fairly well balanced up. Attention is next 
given to the location of the excentric. The points of cut-off, re- 
lease and compression will show whether the angular advance is 
too large or too small, and the readjustment is made accordingly. 
A change in the angle of the excentric for the purpose of adjust- 
ing any one item is moreover liable to disturb other items in 
such way as to require a readjustment of the lap, and is there- 
fore to be avoided unless considered necessary. Thus if the 
cut-off is too late, for example, and the excentric is turned so as 
to increase the angular advance, the cut-off will be made 
earlier and the exhaust and compression as well, while the 
lead will be increased. If the change necessary to adjust the 
cut-off produces too great a disturbance in the exhaust and com- 
pression, or if in general a suitable and satisfactory arrange- 
ment of events cannot be reached by adjustment of the excentric 
and valve rod only, it means that the lap is at fault or perhaps 
the throw of excentric. Change in the lap can of course, only 
be effected by removing the valves and cutting them down if it 
is to be decreased, or fitting a new valve or head if it is to be 
increased. Similarly change in the throw of the excentric can 
only be effected by a removal of the old and fitting a new one 
of proper throw. 






[i] Descriptive. 

An indicator card is a diagram showing for each point of 
the stroke in both directions the steam pressure on the piston. 
Thus Fig. 250 represents an indicator card showing the steam 
pressure above the piston, say, for both the down and up 
strokes. RS is the line of zero pressure from which all pressures 
are measured upward according to the scale of the diagram. 
This is called the absolute pressure line. A is the beginning oi 
the down stroke, B the point of cut-off, C the point of exhaust 

Fig. 250. Indicator Card. 

opening, and D the end of the stroke. The line AB is called the 
steam line and shows the steam pressure on the upper side of the 
piston from the beginning of the stroke to cut-off. The line BC 
is called the expansion line and shows the decreasing values of 
the pressure during that part of the stroke. At C the exhaust 
opens and the pressure drops suddenly as shown by CD. For the 
return or up stroke, D is the beginning, E the point of exhaust 
closure or beginning of compression above the piston, and F the 


point of steam opening just before the beginning of the next 
down stroke. CDE is the exhaust line and shows the nearly con- 
stant pressure during this period. EF is the compression line 
and shows the increasing pressures on the return stroke after the 
closure of the exhaust valve. FA is the admission line and 
shows the sharp jump upward as the steam is opened again just 
before the beginning of the next down stroke. The line PQ 
drawn when the space below the indicator piston is shut off 
from the engine cylinder and connected to the air is called the 
atmospheric line. The distance PR between RS and PQ thus 
represents the pressure of the atmosphere, 14.7 pounds per 
square inch, its length depending of course on the scale of the 
diagram. Thus for the down stroke the varying pressures on 
the top of the piston are shown by the varying distances from 
RS to ABCD, while for the up stroke the pressures on the 

Fig. 251. Pair of Indicator Cards. 

same side of the piston are shown by the distances from RS to 

There will be, of course, a similar diagram for .the head end 
of the cylinder showing the pressures below the piston for both 
the up and down strokes in the same manner as for the diagram 
described. Such a pair of diagrams taken from actual practice 
is shown in Fig. 251. 

Let us now compare the cards of Figs. 250, 251 with Fig. 252, 
the latter showing a so-called ideal card ; that is, a card which 
would be given if the valves opened and closed instantaneously, 
if when closed they were tight against all leakage, if there were 
no loss of pressure due to friction of steam in the passage, and 
if the expansion and compression lines were equilateral hyper- 
bolas. Instead of these conditions the valves open and close 
gradually, even when closed there may be some leakage, there 
is always some loss of pressure due to friction or resistance to 
the flow of steam, especially through a gradually closing or 


opening port and the expansion and compression lines are not 
true hyperbolas. Added to these we have the inertia of the indi- 
cator piston which prevents it from following with absolute ex- 
actness all the variations of pressure as they occur. 

As a result of these various causes the actual engine and in- 
dicator give us the diagrams of Figs. 250 and 251 rather than 
such as Fig. 252. The gradual opening and closure of the valve 
rounds off the various corners, while the steam line instead of 
being horizontal, droops somewhat, due to the loss of pressure 
through the ports and passages. The piston, of course, moves 
faster as it approaches mid-stroke and hence the steam must 
flow in at an increasing velocity to fill up the space behind the 
advancing piston. The higher the velocity the greater the loss 
of pressure, and hence there is a continual slope down from the 
beginning of the stroke as shown in Fig. 251 and often to a far 

Fig. 252. Ideal Indicator Card. 

more pronounced degree. The actual point of cut-off is also 
not always easy to locate, rounded off as it is by the gradual 
closure of the valve. We may, however, properly consider that 
the point of actual final closure is where the curve changes di- 
rection of curvature, that is, from convex to concave, as at or 
near P. Fig. 251. It is sometimes considered that the point of 
equivalent cut-off is more nearly obtained by continuing the 
curve back as shown by the dotted line to Q and supposing 
a sharp cut-off at this point. The result would then be an ex- 
pansion line from Q similar to that which is obtained by the 
gradual closure in the actual case. 

The steam engine indicator diagram is valuable for two 
chief purposes. 

(a) It enables us to judge of the operation of the valve by 
noting the various events, steam opening and closure, the loca- 


tion relative to that of the piston, the resulting piston pressure, 
and to answer various questions relative to the general problem 
of the distribution of steam to the cylinder. 

(b) It enables us to answer all questions which depend on 
the amount and distribution of steam pressure on the piston and 
thus to determine the mean pressure, and knowing the revolu- 
tions to find the indicated horse power; also the turning effort 
at the various points of the revolution, and the mean effort for 
the entire revolution. 

[a] The Indicator Card and the Operation of the Valve Gear. 
We will now consider briefly the more important derange- 
ments which may be met with in the valve gear, and the re- 
sults as shown by the indicator card. 

Fig. 253. Indicator Cards with Angular Advance too large. 

(i) E.vcentric too far from a line at right angles to the crank ; 
that is, angular advance d too large (Sec. 47 [i]). 

Results: Cut-off too early, steam-lead large, exhaust open- 
ing and closure early. In short, the whole round of events is 
ahead of time. See Fig. 253. 

Fig. 254. Indicator Card with Angular Advance too small. 

(2) E.vcentric too near a line at right angles to the crank ; 
that is, angular advance d too small (Sec. 47 [i]). 

Results: Cut-off late, steam lead small or even negative, 
compression small, steam opening late, exhaust opening and 
closure late. In short, the whole round of events is behind time. 
See Fig. 254. 


(3) Steam lap too large. 

Results: Cut-off early, steam opening late and lead small or 
even negative, port opening small with a probable wire draw- 
ing of the steam, and drop of pressure on steam line. See Fig. 


Fig. 255. Indicator Card with Steam Lao too large. 

Fig. 256. Indicator Card with Steam Lap too small. 

(4) Steam lap too small. 

Results: Cut-off late, steam opening early and lead large, 
port opening large. See Fig. 256. 

(5) Exhaust lap too large. 

Results: Exhaust closure early and compression large, ex- 
haust opening late and exhaust lead small. See Fig. 257. 

Fig. 257. Indicator Card with Exhaust Lap too large. 

Fig. 258. Indicator Card with Exhaust Lap too small. 

(6) Exhaust lap too small. 

Results: Exhaust closure late and compression small, ex- 
haust opening early. See Fig. 258. 


(7) Compression excessive. 

Results: The pressure in the cylinder may be carried above 
that in the valve chest before the steam valve opens, thus form- 
ing a loop as shown in Fig. 253. This may be due to either (i) 
or (5) above. 

(8) Expansion Excessive. 

Results: The pressure in the cylinder may fall below that 
in the next receiver or exhaust space beyond, thus forming a 
loop as shown in Fig. 259. 

Fig. 259. Indicator Card with Excessive Expansion. 

(9) Valve Stem too long. 

Results: This means that the middle of the stroke of the 
valve is placed too high relative to the ports. The results for 
an outside valve will be to give too much steam-lap on top and 
exhaust lap on the bottom, and too little steam lap on the bot- 
tom and exhaust lap on top. Hence we shall have : 

Steam opening in top late and small and cut off early. 

Steam opening on bottom early and full, and cut off late. 

Exhaust opening on top early and full and closure late. 

Exhaust opening on bottom late and small and closure 
early. See Fig. 260. 

Fig. 260. Indicator Card with Valve Stem too long or too short. 

(10) Valve Stem too short. 

Results: Similar to those for (9) but oppositely related to 
the ends of the cylinder. 


To these we may also add the following. 

(n) Leaky piston or piston rod stuffing-box. 

Results: The expansion line will be steeper than it should be. 
The compression line may also flatten off somewhat near the 

(12) Port openings or Passages too small. 

Results: Wire drawing or loss of pressure on the steam line 
and rise of pressure on the exhaust line. See Figs. 251, 255. 

It will be noted in the above that different causes may pro- 
duce similar results, so that in interpreting a given set of cards 
caution must be used in working back from result to probable 
cause and remedy. This operation may be aided by the follow- 
ing general hints. 

It will be noted that the general effect of a valve-stem too 
long or too short is to effect the two ends of the cylinder in op- 

Fig. 261. Indicator Card Showing Combination Effect. 

posite directions, thus giving the cards the appearance of hav- 
ing been pushed over in one direction or the other as in Fig. 260. 
On the other hand, if the valve-stem is of proper length but the 
excentric is- improperly set the results will be of the same kind in 
both ends of the cylinder as shown in Fig. 253. Various com- 
binations of these may exist in the same engine. Thus a pair 
of cards as shown in Fig. 261 indicates an incorrect length 
of valve-stem, an incorrect adjustment of the laps, with perhaps 
too large an angular advance. The combination nearly corrects 
certain difficulties and makes others still worse. 

Various special features may combine to make the so-called 
"freak" cards, but we shall not examine this part of the subject 
further as such freaks are of rare occurrence, and a careful study 
of the results of the various single derangements as given above 
in (i) to (12) will usually be sufficient to show the nature of the 


[3] Working Up Indicator Cards for Power. 
From the principles of mechanics we know that work is the 
result of a force or effort acting through a distance, and is 
measured by the product of the force in pounds by the distance 
in feet. This gives the measure of the work in foot-pounds. 
Power measures the capacity to perform a certain amount of 
work in a given time. The common unit is the horse power, 
which is 33,000 foot-pounds of work done in one minute of time. 
Hence to find the power of an engine we have two chief steps : 

(1) To find the foot-pounds of work done per minute. 

(2) To reduce this to horse power by dividing by 33,000. 

It may be noted here that the term Indicated Horse Power 
means simply the horse power as: determined from the indicator 

Now by definition the foot pounds per minute for the steam 
engine will be the product of the acting force multiplied by 
the distance through which it acts in one minute. The acting 
force equals the mean load on the piston, and this equals th^ 
mean effective pressure per square inch multiplied by the area 
in square inches. The distance acted through per minute must 
be measured in feet, and equals twice the stroke multiplied by 
the number of revolutions per minute. 

Let p = mean effective pressure in pounds per square inch ; 
A = area of piston in square inches ; L = length of stroke in 
feet ; N = revolutions per minute. Then pA acting force or 
mean total load on the piston measured in pounds, and 2LN 
distance moved per minute in feet = piston speed. Hence 
foot-pounds of work per minute equals product (pA) X (2LN) 
or what is the same thing 2pLAN. Hence we have the formula : 

2 pLAN 
Horse power = - 


This is the usual formula for finding the indicated horse 
power, and is commonly employed for working up indicator 
cards for this purpose. 

The reasons for measuring L in feet and A in square inches 
will be readily seen from the following considerations. Work 
is composed of two factors, the force factor and the distance 
factor. The first must be measured in pounds and the second 
in feet. The product pA is the force factor, and since / is 
usually measured in pounds per square inch, A must be measured 


in square inches in order that pA may be the total mean load 
in pounds. The product 2LN is the distance factor, and hence 
2.L the distance traveled per revolution must be measured in 
feet, in order that 2.LN may be the distance traveled per minute 
measured in feet. The product (pA) X (2ZJV) or 2pLAN will 
then give the work measured in foot-pounds as we have seen 

We will now give by rule the operations necessary to find 
the indicated horse power, as expressed by the formula above. 

Rule Multiply together the mean effective pressure in 
pounds per square inch by the length of the stroke in feet, 
and this product by the area of the piston in square inches, anrl 
this product by the number of revolutions per minute, and this 
product by 2, and then divide the final product by 33,000. The 
quotient will give the indicated horse power. 

Fig. 262. Mean Effective Pressure from Indicator Card. 

Now the various factors which enter into either the formula 
or rule for horse power, the length of stroke and area of the 
piston come from the dimensions of the engine, and the revolu- 
tions per minute from the counter, or by actually counting them, 
watch in hand. There remains the mean effective pressure p 
which must be found from the indicator cards, and to this part 
of the operation we now turn. 

The mean pressure for a single card such as Fig. 262 gives 
simply the mean of the pressure in one end of the cylinder, say 
the top. To obtain this mean pressure we may proceed in a 
number of different ways. Fundamentally the mean of such a 
series of pressures as given by the indicator card, is found by 
dividing the area of the card by the length. This gives the 
side of a rectangle which would have the same area as 
the card. Thus in Fig. 262 if the rectangle ABCD has the 
same area as the card, then the side A D of the rectangle 
is the. mean height of the card, and to the proper scale 



will give the mean pressure desired. Hence any method 
which will give the area of the card may be used for 
rinding a mean height, and hence a mean pressure. In 
Part II., Sec. 9 [15] are given various rules and methods for 
finding the measure of an irregular area, illustrated by the ex- 
ample of an indicator card, and any of these may be used as 
there explained. The method most commonly used is to meas- 
ure the ordinates on the dotted lines as in the figure there shown, 
take their sum, and divide by their number, 10. This multiplied 
by the scale of the indicator spring will give the mean pressure 
desired. The simplest method of locating the intervals for these 
dotted ordinates is that explained in Part II., Sec. 10 [4]. To 
carry this out we proceed as follows : 

Let the card be represented in Fig. 263, then draw the lines at 
the ends as shown, perpendicular to the atmospheric line OA 

Fig. 263. Subdivision of Indicator Card for Obtaining Mean Ordinate. 

and tangent to the card, thus fixing its length. Then lay off 
the line OB at an angle and on OB lay off first a half divi- 
sion Oi, then nine whole divisions, and then a half division as 
shown. The divisions may be taken from y to y 2 inch in ac- 
cordance with the length of the card. Then drawing a line 
from B to A and other parallel lines from the points of division 
on OB to OA, the locations for the ordinates are determined, 
and they may be drawn as shown. Where a large number of 
cards are to be worked up in this way, time will be saved by 
the use of a form of template or pattern for locating these points. 
Such an implement is shown in Fig. 264 and consists of a piece 
of hard wood with small steel points set in to the edge, spaced 
according to the lay out of points along OB Fig. 263. The 
distance between the extreme points is somewhat greater than 
the length of the longest card likely to be met with. Instead of 



steel points set in a block of wood, a thin plate of steel may be 
cut out and filed up so as to leave the points projecting at the 
desired intervals. In using this device it is simply necessary to 
draw lines tangent to the ends of the card as shown, and then 
to place one end of the template on one boundary line PR at 
any convenient point as P, and swing it to such an angle as 
will just bring the other end Q to the other line QS. The 
template is then pressed down so as to mark the paper with the 
points, and lines parallel to those at the ends are drawn through 
the points thus marked, as shown by the lines of the figure. 
In this way the ordinates spaced in the manner desired may be 
rapidly laid out and drawn in. 

For summing the ordinates the method by the use of a strip 
of paper as explained in Part II., Sec. 9 [15] may be recom- 


i i 

Fig. 264. Subdivision of Indicator Card for Obtaining Mean Ordinate. 

mended as the simplest, quickest and most satisfactory available 
for the purpose. 

Having thus in one way or another found the mean effective 
pressure for one card, the other one of the pair is taken in like 
manner, thus giving tlie mean effective for the other end of the 
cylinder or other stroke. These two values may then be 
averaged, and the result taken as the mean effective pressure 
for the revolution, thus furnishing the final factor p required in 
the formula or rule for horse power. 

It must be noted that this operation is slightly in error by 
reason of the difference in area between the upper and lower 
sides of the piston. On the upper side the entire area is effective 


while on the lower side the piston rod takes out a small area in 
the center. To take account of this, we may compute the I.H.P. 
for each end of the cylinder separately. To this end we take 
each card by itself, say the head end first, and find the mean. 
effective pressure which we may denote by / x . Let the entire 
piston area he A l . Then as before the mean load or average 
acting force is the product of the two, />, A r The distance acted 
through is L for each down stroke, and the number of down 
strokes per minute is equal to the number of revolutions A r . 
Hence the distance per minute for the down strokes is LN and 
the I.H.P. for this end of the cylinder will be: 


33000 33000 

In a similar manner we then find the mean effective pres- 
sure for the bottom or crank end of the cylinder which we may 
call pi. Then taking from Ai the area of the piston rod, we have 
the effective area of the bottom of the piston which we may 
call At. Then similarly as in the head end we have for the 
I.H.P. in the crank end, 

H --- P* A LX ^P~- LA - N 

33000 33030 

The total I.H.P. will then be the sum of these for the two 
strokes up and down, or : 

IH.P. ^ H,-\ H,= 

For illustration see example (7) below. 

Mean Effective Pressure by the Aid of the Planimctcr. 

The planimeter, an instrument for measuring areas, is also 
frequently used for working up indicator cards, and where the 
number is large will be found of great service. Such instru- 
ments may be. obtained of most makers of indicators or of 
dealers in mathematical instruments. General directions for 
their use will accompany them. The following hints may be 
given for their use with indicator cards. 

Where the instrument has an adjustable bar it should be 
sel so ns to read the area in square inches. \Yhere the bar is 
not adjustable the instrument is usually already set to read in 
terms of this unit.- The order of procedure is then as follows : 

(i) Draw lines at the ends of the card at right angles to 
the atmospheric line so as to be able to determine its length. 


(2) Place the instrument and card in a suitable position, 
and read the record wheel, putting down the result, say 3.26 as 
below : 

Readings. Differences. Average. 

First 3.26 

Second 7.08 3.82 

Third 10.92 3.84 3.83 

(3) Then trace around the contour, usually in the direction 
with the hands of a watch for a second reading greater than the 
first, and come back carefully to the starting point. Then read 
again and set down the result, say 7.08 below the first as shown. 

(4) Then repeat, tracing around as before, read and set 
down the result, say 10.92, below the others as shown. In mak- 
ing the last reading it will be noted that on the instrument itself 
we might be able to read only 0.92, but the increase upward 
from 3 to 7 shows that the wheel has passed the starting point 
and begun again, so that we must add the ten and write 10/12. 

(5) We then take the difference of the readings, the first from 
the second and the second from the third and set down as shown, 
and then average these two numbers, thus finding in the present 
case 3.83 for the area in square inches. The reason for going 
around the area twice is to have two measurements, so that each 
will give a check on the other. If they differ widely an error 
somewhere is certain, and they must be repeated, while if nearly 
the same, as in the case given above, the error is no more than 
must be expected with such means, and the average may be 
taken as the value of the area desired. 

(6) We next divide the area by the length of the card. Thus 
suppose in the case in hand that the length is 4.2 inches. Then 
3.83 -r- 4.2 = .912 inches. This is the mean ordinate or mean 
height of the card in inches. 

(7) We next multiply by the scale of the indicator spring 
and thus find the mean effective pressure desired. Thus sup- 
pose the spring to be 60 pounds to the inch. Then 60 X .912 = 
54.72 pounds. This is then the mean effective pressure for the 
stroke as given by the card thus measured. 

We then proceed similarly with the other card, and use the 
results for the determination of horse power in the manner al- 
ready explained. 


Illustrative Examples. 

(1) The area of an indicator card is 2.87 sq. in. and its 
length is 3.8 in. What is the mean height? 

Solution: 2.87 -4- 3.8 = .755 in. 

(2) The scale of the indicator spring is 40 Ibs. per inch. 
What is the m.e.p.l* 

Solution: .755 X 40 = 30.2 Ibs. 

(3) The ordinates measured in inches taken from an indi- 
cator card divided up as in Fig. 263 are as follows : 

.91, 1.30, 1.44, 1.40, 1.35, 1.20, .95, .80, .70, .25, and the 
scale of the indicator spring is 60 Ibs. per inch. Find the m.e.p. 

Solution : Adding the lengths as given, we have for the sum 
10.40. Hence dividing by 10 we have for the mean ordinate 
10.40 -f- 10 1.04. Hence the m.c.p. is 60 X 1.04 = 62.4 Ib. 

(4) The total length between marks on a strip of paper 
used to measure the ordinates as described in Part II., Sec. 
9 [15] is found to be 6.3 in. The scale of the spring is 20 Ib. 
Find the m.c.p. 

Solution: 6.3 -f- 10 = .63 in. = mean height, and 
.63 X 20 = 13.6 Ib. = m.e.p. 

(5) Given an indicator card with ordinates spaced as in Fig. 
263. The pressures measured by a scale corresponding to the 
indicator spring are as follows : 

18, 26, 28.4, 27.8, 27, 24.2, 19, 16, 14.3, 7.2. Find the m.e.p. 

Solution: We add the pressures and find the sum 207.9. 
Divide this by 10 and we have 20.79 or 2o -8 Ib. as the value of 
the m.e.p. 

(6) From the two cards of a pair the values of the m.e.p. are 
found to be 28.6 for one end and 32.2 for the other. The piston 
area is 1,213 sc l- m -> the stroke 39 in. and the revolutions 102. 
Find the I.H.P. neglecting the effect due to the area of piston 

Solution: The m.e.p. for the whole revolution is the mean 
of the values for the two ends or m.c.p. (28.6 X 32.2) -f- 2 

= 304- 

Then stroke in feet = 39 -^ 12 = 3.25. 

Then I.H.'p. : = 2 X 3 ' 4 X 3 ' 25 x I213 X I02 


* This abbreviation is often used for the term mean effective pressure. 


Multiplying out the factors of the numerator and dividing by 
the denominator we find I.H.P. = 741 Ans. 
(7) Given the following : 

Diam. of cylinder = 24 in. 
. Diam. of piston rod = 5 in. 
m.e.p. from head end or pi = 63.4 Ib. 
m.e.p. from crank end or p* = 58.8 Ib. 
Stroke = 36 in. 
Revolutions no. 

Find the I.H.P. both with and without the allowance for the 
area of piston rod. 

Area of 24 inch piston or Ai = 452.4 sq. in. 
Area of 5 inch piston rod or a = 19.6 sq. in. 
Effective area of lower side of piston = difference, or A* 
= 432.8 sq. in. 

Then neglecting the effect of the rod we should say : 
m.e.p. = (63.4 + 58.8) -r- 2 = 61.1 

2 X 61.1 X 3 X 45 2 -4 X no 

Working this out we find: I.H.P. = 552.8. 
Taking account of the piston rod area we have for the head 

H = 6 3-4 X 3 X 452.4 X no =2g6 g 


For the crank end : 

= 58.8x3^433-8X110 = 

Adding we have : 
H = 541.3. 

There is thus seen to be in this case a difference of 11.5 
horse power, constituting an error by the first method of some 
considerable amount. It is readily seen that this error will be 
relatively less the larger the cylinder, especially in the cylinders 
of a multiple expansion engine. Thus in the case given which 
was for the H.P. cylinder of a triple expansion engine the error 
is 11.5 horse power, or about 2 per cent. For the I. P. cylinder 
the error would be not far from 4.5 horse power or about .8 per 
cent., while for the L.P. cylinder it would be perhaps two horse 
power or about .3 per cent. This would give a resultant error 

and J.H.P = /'X6..ix 3 X45'.4Xiiox 
V 33000 / 



of about i per cent, for the engine as a whole. While these 
figures would vary with particular circumstances, they will serve 
to illustrate the nature of the error, and the methods given show 
how to avoid it when so desired. 

[4] Combined Indicator Cards. 

The cards taken from the various cylinders of a multiple 
expansion engine, as for example those of Fig. 265, may be 




Fig. 265. Set of Indicator Cards from Triple Expansion Engine. 

combined in such a manner as to show very instructively the 
continuous history of the expansion of the steam, that is the 
continuous relation between volume and pressure as the steam 


passes through the engine. To effect this combination it is 
necessary to lay down the various cards in one diagram and all 
to the same scale of volume and pressure. The details of the 
operation may be sketched out in the following steps : 

(1) In Fig. 266 take the two lines at right angles, OX and 
OY, the former as an axis of volume and the latter as an axis of 

(2) Determine in cubic feet for each of the cylinders the 
volume of the/ clearance (Sec. 67), and the volume swept by the 

(3) Lay off the lines AB, CD, EF at such distances from 
OY as to represent respectively the clearance volume in the 

Marine Engineering Q X. 

Fig. 266. Combined Cards from Triple Expansion Engine. 

H.P., I. P., and L.P. cylinders, taking care to select the scale of 
volume such that the L.P. volume plus its clearance as measured 
between the lines OY and GK will come within the desired limits 
of the diagram. 

(4) Lay off on each card the line of zero pressure or the 
perfect vacuum line, as shown by OX in the small diagram A. 

(5) Take next the H.P. card as at A for example, and select 
any point such as P. Measure in any convenient units the dis- 
tances MP and MN : multiply the volume of the cylinder by the 
former and divide by the latter. This will give the volume 


swept in the H.P. cylinder from the beginning of the stroke to 
the point P. 

(6) The corresponding point P of the combined diagram 
is then found by measuring from AB a distance HP representing 
this volume, and from OX a distance JP representing the pres- 
sure PO on the card. This will give the point P, and other 
points are found in a similar manner, as many as may be needed 
to determine the form of the card as shown. It is to be especially 
noted that the H.P. card of the combined set is the same as that 
at \ but drawn simply with different scales, and therefore more 
or less distorted in appearance. 

(7) The points necessary to determine the other cards of 
the combination ae found in a precisely similar manner, re- 
membering that in each case volume is measured from the clear- 
ance line CD or EF, while the pressure must be measured from 
the line of zero pressure for the card and laid off from the cor- 
responding line OX of the combined set. 

This diagram shows the general manner in which the steam 
expands on its way through the engine. An expansion line 
PQ shows the general law of expansion as a continuous 

PR is an ideal expansion line laid down as a hyperbola, all 
points in the curve corresponding to the condition that the 
product of volume by pressure shall be constant, or in symbols, 
pv Constant. This shows the result of the so-called true 
hyperbolic expansion law, and as appears from the diagram, the 
actual expansion line is somewhat below this ideal line. 

The equation to the actual expansion line may be expressed 
in the form pv" = Constant, where n is an exponent having 
values usually lying between 1.15 and 1.2. The equation pv 1 - 18 
may be taken as very commonly representing this line in good 
average practice. The extent to which the area bounded by 
the line PR and the clearance lines on the left is well filled in. is 
an indication of the degree to which the performance of the 
actual engine approaches that of an engine having true hyper- 
bolic expansion and with indicator cards as shown in Fig. 252. 
The relation between the actual engine and such an ideal case 
is usually expressed by a percentage factor known as the "card 
factor." For good practice with triple expansion engines, this 
factor will be found from .60 to .70. With quadruple expansion 
engines representative values are found from .55 to .60. 


The diagrams of figures 265 are reproduced from an actual 
case and may be considered as representing good modern 
practice in general character and form. 

At this point reference may be made to the effect on the 
distribution of power in a compound or multiple expansion 
engine, of linking up or cutting off earlier in the intermediate 
or low pressure cylinders. Taking first the case of a compound, 
linking up or shortening the cut-off on the L. P. cylinder will 
increase the power in this cylinder and decrease it in the high. 
This result at first sight seems contradictory to common ex- 
perience, because in a single cylinder we are accustomed to 
associate an earlier cut-off with decrease of power. In the case 
of the compound, however, cutting off earlier in the L.P. cylin- 
der gives a higher back pressure in the H.P. cylinder and a 
consequently higher initial pressure in the L.P. cylinder, and 
thus results in an actual addition to the L.P. indicator card area 
instead of a decrease as in the case of a single cylinder. At the 
same time the area of the H.P. card will be reduced and the 
power developed in this cylinder will be decreased correspond- 
ingly. Similarly for a multiple expansion engine and in general, 
cutting off earlier in any of the cylinders beyond the first or 
H.P. will result in an increased back pressure for the next pre- 
ceding cylinder, and in a higher initial pressure for the cylinder 
itself, and thus in an actual addition to the area of the indicator 
card and a corresponding subtraction from the area of the card 
for the cylinder preceding. 

Thus in Fig. 266, cutting off earlier in the L.P. cylinder 
will result in raising the upper line of the L.P. and lower line 
of the I. P. cards, and thus in increasing the area of the former 
and decreasing that of the latter. In like manner cutting off 
earlier in the intermediate cylinder will result in raising the 
upper line of the I. P. and lower line of the H.P. cards and thus 
in increasing the area of the former and decreasing that of 
the latter. In like manner cutting off later in any cylinder be- 
yond the H.P. will result in similar changes but in the opposite 
direction. Thus a later cut-off in the I. P. cylinder will decrease 
the power developed in that cylinder, and increase the power 
developed in the H.P. cylinder. It thus results that a combi- 
nation of changes such as a later cut-off in the I. P. cylinder and 
earlier cut-off in the L.P. will both tend to decrease the power 
developed in the I. P. ; while an earlier cut-off in the L.P. cylinder 


and a later cut-off in the H.P. will both tend toward an increase 
of power developed in the I. P. 


[i] Descriptive. 

The indicator card has already been described in Sec. 55. 
It is the purpose of the indicator to draw this card. It must 
therefore provide for the proper combination of these move- 
ments, (i) A movement in step with the piston and propor- 
tional to it in amount so that all horizontal distances on the card 
shall bear a constant proportion to the corresponding parts of 
the stroke. (2) A movement at right angles to that in (i) and 
in direct proportion to the pressure per square inch on the pis- 
ton in the end of the cylinder to which the indicator is con- 

The combination of these movements will then result in a 
diagram such as those shown in Sec. 55, and giving at each 
point of the stroke the pressure on the piston as desired, the 
upper line showing the pressure which urges the piston for- 
ward on one stroke and the lower line the pressure which re- 
sists its movement backward on, the return stroke. 

In Fig. 267 a modern indicator is shown. A is a drum to 
which the paper is attached by means of the clips as shown. 
This drum is given a motion back and forth about its axis by 
means of a connection with the crosshead through the so-called 
"reducing motion." By this means the drum is given a motion 
of some three to five inches in extent, just in step with the 
motion of the piston and proportional to it in amount. B is 
the indicator cylinder or barrel connecting with the end of the 
engine cylinder from which the card is to be taken. Within 
the cylinder, as shown, is a piston with a coiled steel spring 
above, resisting pressure from below the piston upward. To 
the piston rod is attached a linkage carrying at the end of 
the arm P the pencil point which is to trace the diagram upon 
the paper carried by the drum. The connection between the 
linkage and the piston rod is such that the former may be swung 
freely about the cylinder upon a ring to which it is attached. 
The pencil may thus be brought into contact with the paper on 
the drum or withdrawn from it as desired. An adjustable screw 
stop is provided, and so arranged as to arrest the movement 



of the pencil motion when swung around by the hand, and thus 
allow only light contact between the pencil point and the paper. 
In some cases a brass point is used instead of a pencil, the 
cards being of paper specially prepared so that the brass will 
leave a black mark upon it. Such points are strong and require 
no sharpening except at long intervals. 

The object of the linkage which, forms the pencil motion 
is to magnify the movement of the indicator piston, and thus 

Marine Engineering 

Fig. 267. Steam Engine Indicator. 

to allow the use of stiff springs with a corresponding small 
movement of spring and piston. With high revolutions es- 
pecially, this is found necessary in order to reduce as far as 
possible the disturbance in the diagram due to the inertia of 
the moving parts of the indicator. The linkage is thus a form 
of multiplying motion, or a reducing motion reversed, and it 


should give to the pencil a movement exactly proportional to 
that of the piston, but 3 to 5 times greater as may be desired. 

The relation between the pressure per square inch and the 
actual movement at the pencil point fixes the so-called scale of 
the spring. This depends also on the actual area of the indi- 
cator piston, which is, however, usually about one-half square 
inch. Thus a 40 pound spring means a spring such that a pres- 
sure of 40 pounds per square inch on the indicator piston, or say 
an actual load of 20 pounds, will produce a movement of one 
inch at the pencil point. 

By means then of the piston, spring and linkage, the second 
of the necessary movements as mentioned above is thus pro- 

Returning to the drum the first of the motions above noted 
is obtained by some form of reducing motion as described be- 
low. The connection between the drum and the reducing mo- 
tion is usually made by means of a cord C wrapped around 
a groove in the base as shown. The cord thus serves to pull 
the drum around in one direction while the return stroke is 
made by means of a coiled spring in tfie base. This spring op- 
poses the motion given by the cord, and is therefore coiled up 
during the forward stroke. As soon, however, as the pull of the 
cord ceases the spring takes charge and uncoiling carries the 
drum in the reverse direction as fast as the cord will allow, thus 
keeping the latter taut and insuring the motion of the drum in 
step with the main piston in both directions as accurately as the 
form of reducing motion may determine. 

A separate indicator may be provided for each end of the 
cylinder, or by suitable pipe connections and a three way cock, 
one indicator may be made to serve for both ends. In any 
case the cock which shuts off the indicator must be so arranged 
that when shut off from the cylinder the space below the piston 
will be connected to the outside air. The piston with equal air 
pressure on both sides will then come to a position of equilib- 
rium, and the atmospheric line may be drawn. 

[2] Reducing Motions. 

The purpose of the reducing motion has already been 
stated. There are many different ways in which the desired 
movement may be given to the drum, some of them accurate 
in geometrical principle and some only approximate. 



One of the most common is by means of links, levers and 
bell-cranks. The simplest of such forms is shown in Fig. 268. 
A is a pin attached to the crosshead. AB is a short link con- 
necting the crosshead to a lever BD pivoted at C. The point 
D will then move in step and nearly in constant proportion to 
the piston, and from D the motion for the drum may be taken, 
either by a cord direct, or from the end E of a rod DE moving 
as shown. In such cases the cord should run in continuation 
of the line DE and not off at an angle as EF or DH. As a 
general rule in all such cases, the reducing motion should be 
so adjusted that the cord part should not undergo changes of 
angular direction, or at least such changes should be made as 
small as possible. Thus in Fig. 269 suppose the point from 

Fig. 268. Reducing Motion. 


B" Marine n 


Fig. 269. Reducing Motion. 

which the. motion is taken to move through a path AB, and 
the indicator guide pulley to be at P. Then at one extreme the 
cord will be represented by PA, and at the other by PB. Such a 
change in the angular direction of the cord relative to the line 
of motion AB will result in error, and should be avoided by 
bringing P over AB or AB under P. It is not necessary that 
the motion of the point E Fig. 268 should be vertical so long 
as the gear is so arranged as to reduce to a minimum all angu- 
lar changes in cords and connecting links. Thus the arrange- 
ment of Fig. 270, while containing a large number of joints and 
parts, may be as nearly correct as the simpler form of Fig. 268. 
Instead of taking the motion direct from D a link DG con- 
nects this point with a bell-crank GHI pivoted at H. Then a 



second link 1J connects this to a second bell-crank JKL and a 
rod LE guided at M gives a point E from which the motion- 
may be taken, or if more convenient the rod LE may be dis- 
pensed with and the motion taken from L direct. Such a com- 
plication of gear is of course not desirable, and the arrangement 
is shown simply as an illustration of a combination of links and 
bell-cranks which would still give the motion required. 

A - 


Fig. 270. Reducing Motion. 

All such forms of reducing motion are approximate and not 
geometrically exact. The error is, however, in most cases small 
and is usually neglected, though if desired its nature and extent 
may be investigated by a suitable geometrical analysis of the 

Instead of taking the motion from the crosshead by means 
of a short link as AB Fig. 268, a lever BD Fig. 271 is some- 
times provided, having a forked end and pivoted at C or D. A 
pin on the crosshead working in the slot or forked end gives the 
to and fro motion to the lever, while from D or C the desired 
motion is taken. 

Fig. 271. Reducing Motion. 

Instead of attaching the cord direct to C for example, a 
sector of wood PQD with center at D is attached to the arm, 
and the string is led off from the face of the sector. Such a 
sector may also be employed with the arrangements shown in 
Figs. 268 and 270. None of these motions is geometrically 

A form of pantograph consisting of jointed rods as shown 
in Fig. 272, may sometimes be used when there is room for it 



to work freely. A is attached to the crosshead and D or E is 
the fixed pivot. Then the other point E or D will provide a 
motion for the indicator drttm which is geometrically exact. 
Here again however, the string should be so led that its angu- 
larity will not vary. 

Instead of this arrangement of links the so-called lazy-tongs 
as shown in Fig. 273 is sometimes employed. This is also 
geometrically exact, and is in fact an equivalent to the panto- 
graph in Fig. 272, without requiring quite as much room. 

Various combinations of pulleys may also be used, as illus- 
trated in Fig. 274. AB is an arm projecting from the crosshead 
and moving with it. To the end B of this arm is attached a 
cord wrapping around a light pulley P. Q is a smaller pulley on 

Marine UnjiMerinj 

Fig. 272. Pantograph Reducing Motion. 

Fig. 273. Lazy Tongs Reducing Motion. 

the same axis and moving with P. Wrapped on this is a cord 
CD, which may be led off in various directions to the indicator as 
shown by CD, CD^ CD L> . This gear is geometrically exact. 

Various other forms of reducing motion are also to be met 
with, but those described will be suffcient to show the forms 
most commonly available for marine practice. 

[3] Taking an Indicator Card. 

The instrument should first be examined and put into proper 
condition and adjustment. This should include the following 
points : 

(i) The joints should all work freely, but without lost 



(2) The piston should not bind nor should it be so loosely 
fitted as to allow serious leakage. A slight leakage is, however, 
better than too snug a fit. 

(3) The working surfaces of the barrel and piston should 
be carefully wiped and oiled. This should be repeated from time 
to time when a series of cards is being taken. The joints of the 

Fig. 274. Reducing Motion. 

pencil motion should also be lubricated with clock oil as often as 

may be required. 

(4) The pencil points should be sharpened and the screw 

stop so adjusted that the point can rest only lightly on the paper. 
The operation of taking the card itself is briefly as follows : 
The indicator is attached to the cock, a blank card is placed 

on the drum and the cord connection is adjusted so that the 

drum will have the proper stroke without coming against the 

Fig. 275. Putting on an Indicator Card. 

stop at either end. In attaching the blank card the most con- 
venient way will be to bend the sheet of paper around and grasp 
both edges between the thumb and forefinger as at AB in Fig. 
275a. Then slip over the drum and under the clips so that the 
latter will come outside the paper as shown at PQ, b. Then 
slip the paper down into place, pull and adjust so that it fits snug- 


ly, and bend the edges back as in c. The cord is then hooked 
on to the reducing motion and the drum takes up its movement 
with the main piston. The cock is then opened to the end of 
the cylinder from which the diagram is desired, and the pencil 
immediately takes up its motion corresponding to the varying 
pressures of the steam. The indicator piston should be allowed 
to work in this way for a few strokes, or until everything is 
warmed up into working condition. 

When everything is in readiness the pencil motion is moved 
up against the stop so that the pencil resting lightly on the paper 
will trace its path for a complete revolution or longer if de- 
sired. Then remove and shut off the indicator from the cylinder. 
This will connect it with the air, the indicator piston will come 
to equilibrium under atmospheric pressure, and the atmospheric 
line may then be drawn. The drum connection is then un- 
hooked, the paper removed, a fresh one replaced, and the next 
card taken when desired. If one indicator is used for both ends 
of the cylinder, both cards should be taken on the same paper 
with as small an interval between as possible. The cock is 
swung over for one end and the card taken, and then imme- 
diately swung over for the other end and the second card taken 
without loss of time. The cock is then closed off connecting 
the indicator with the air, and the atmospheric line is then 

Each card as it is removed from the indicator should be 
marked with sufficient data to identify it, and make possible its 
use for the purpose intended. This should include at least the 
following items : 

(1) Cylinder. 

(2) End from which card is taken. 

(3) Revolutions. 

(4) Scale of spring. 

(5) If a series of cards is being taken the time and serial 
number should also be set down. 

The various other items usually printed on the back of the 
card may be filled in at a later time as may be convenient. When 
cards from both ends are taken on one paper, we must be able 
to assign each to its proper end of the cylinder. The most cer- 
tain way of determining this is to shut off the connection to one 


end of the cylinder entirely, and then take the card from the 
other end. It will thus appear how the card from this end lies 
on the paper, whether with admission line to the right or left, 
and this will show how to mark the entire series of cards taken 
with the same arrangement of reducing gear, etc. 





[i] Constitution of Matter. 

For the purpose of explaining or discussing the relations be- 
tween matter and the forces of nature, all substances are sup- 
posed to be composed of enormously large numbers of in- 
definitely small parts called molecules, each one of which is 
supposed to be, in fact, the smallest portion of the substance 
which can exhibit its various properties. These molecules are 
furthermore not at rest, but are supposed to be in a state of 
more or less violent agitation or motion. If the motion of each 
molecule is about a fixed center so that they all retain their 
average positions fixed in the body, it is said to be a solid or in 
the solid state. If the motion of the molecules is about centers 
which themselves are free to move about in any direction, so 
that the average position of the molecules is not fixed and the 
body readily changes its form, it is said to be a liquid or in the 
liquid state. If the motion of the molecules is in straight lines 
hither and thither, bound to no center or location, but ever 
striving to fly as far apart as possible, the substance is said to 
be a gas or in the gaseous state. 

In the solid and liquid states the molecules are bound to- 
gether by forces of molecular attraction, so that they tend to 
maintain about the same average distance apart, and thus to fill 
the same volume. Any attempt to change this average distance 
between the molecules and thus to make the volume larger or 
smaller, must deal with these molecular forces. The only prac- 
ticable way of doing this is through the agency of heat, as we 
shall see in the next section. In a gas the forces binding the 
molecules together have been overcome, and the molecules have 


been separated so much further apart that all traces of these 
attractive forces have disappeared, and instead we now have a 
repulsive force acting between the molecules and urging them 
as far apart as the limits of the volume which contains them 
will allow. Due to this property a gas will expand and fill any 
volume, no matter how large, the repulsive force or force of ex- 
pansion, however, becoming weaker as the volume increases 
and the average distance between the molecules becomes greater 
and greater. 

[2] Heat. 

(i) Heat and Its Relation to Matter. We know energy as the 
capacity for doing work. Also the energy of motion is called 
kinetic energy, while the energy of position or location relative 
to a given force, is called potential energy. 

Heat is one of the many forms of energy. It is, in fact, the 
energy of the molecule, and the heat in a body means therefore 
simply the amount of such molecular energy which the body 
possesses. This energy of the molecule is partly kinetic or due 
to its motion, and partly potential due to its position relative to 
the molecular forces which act upon it. The addition of heat 
to a body or its subtraction from a body means therefore the 
addition or subtraction from the energy of its molecule, and 
hence the addition to or subtraction from its store of molecular 
energy. This addition or subtraction of heat is always accom- 
panied by a series of changes in the state or condition of the 

Thus if heat be added to a lump of ice at the melting point 
or 32 Fah. it first melts or changes from a solid to a liquid, 
remaining at the constant temperature of 32 and slightly con- 
tracting in volume meanwhile. If heat is still farther added the 
water grows warmer to the touch and as shown by the ther- 
mometer. At the same time it continues to contract slightly till 
it reaches a temperature of about 39 Fah. and then slowly ex- 
pands. If under atmospheric pressure the increase of tempera- 
ture and volume will continue with the addition of heat until 
the thermometer marks a temperature of 212 Fah. Then the 
further addition of heat occasions no further elevation of tem- 
perature, but instead, a new change of state. The water now 
passes into the state of vapor or steam, the temperature of both 
the water and the vapor formed from it remaining meanwhile 


constantly at the fixed temperature of 212. After the water is 
completely vaporized, if the vapor be inclosed in a chamber of 
fixed volume and heat still further added, it will then be found 
that the pressure and temperature will continue to increase so 
long as additional heat is supplied. If instead the pressure is 
kept constant, the volume and temperature will increase as heat 
is added. If the temperature is kept constant, the pressure will 
fall as the volume is increased. 

From the start then, as more and more heat has been added, 
the water has exhibited successively the three states of matter. 
As ice it is a solid; in its usual state or between 32 and 212 
under atmospheric pressure, it is a liquid; and after it is com- 
pletely vaporized and heat is still further added so as to carry 
the conditions considerably beyond those at which the vapor 
was formed, it becomes a gas. 

We must here explain the difference which may be implied 
in the words gas and vapor. When a substance first changes 
from the liquid to the gaseous state, or while the pressure, 
volume and temperature are near those corresponding to such 
a change, the substance is more strictly called a vapor, or is said 
to be in the vaporous condition. If the substance is in the gaseous 
state but with pressure, volume and temperature conditions far 
removed from those corresponding to the change of state, the 
substance is more generally called a gas. There is no sharp line 
of difference between a vapor and a gas. The former means 
simply a substance in the gaseous state, but at or near the con- 
ditions corresponding to the process of change from one state 
to the other, while the latter means likewise a substance in the 
gaseous state, not far removed from the conditions correspond- 
ing to the process of change of state. 

There are thus two chief kinds of change which the applica- 
tion of heat may produce. 

(a) It may change the temperature of a substance accom- 
panied by a change of pressure or volume or both, but without 
change of state. 

(b) It may produce a change of state as from solid to liquid 
or liquid to vapor, accompanied usually by a change of volume, 
but without change of temperature. If, however, the pressure 
varies during the change of state, then the temperature at which 
the change occurs will also vary, but there will be no change 
of temperature directly accompanying the change of state. In 


consequence, during the change from solid to liquid or vice versa, 
the temperatures of both are the same, and similarly during the 
change from liquid to vapor or vice versa, the temperatures of 
both are the same. 

It must be understood when changes are referred to as 
depending on the addition of heat, that the subtraction of heat 
will produce changes in exactly the opposite direction. Thus 
if the addition of heat causes a body to expand, the subtraction 
will cause it to contract : if the addition causes an increase of 
pressure the subtraction will cause a decrease : if the addition 
causes a change of the state from liquid to vapor, the subtrac- 
tion will cause a change from vapor to liquid, etc. 

(2) Sensible and Latent Heat. Heat which causes a change 
of temperature in a body, as when water is heated and becomes 
hotter to the touch or to the thermometer, is called sensible heat. 
This corresponds to a change in the kinetic energy of the 
molecule, so that increase of velocity of the molecule and in- 
crease of its kinetic energy correspond within the substance to 
the growing hotter to the touch and to increase of temperature 
as observed on the outside. 

Heat which is involved in a change of state but which pro- 
duces no effect on the temperature of the substance (as in the 
melting of ice at 32 or the boiling of water at 212) is called 
latent heat. This corresponds to a change in the potential energy 
of the molecule so that an increase in the average distance 
between the molecules (acquired in opposition to the molecular 
forces acting), and a consequent increase in their potential 
energy corresponds within the substance, to the change of state 
at constant temperature as observed on the outside. 

It must be understood that there is really but one kind of 
heat, and that this division into sensible and latent is only a 
matter of convenience in order to signify the particular energy 
change which is effected within the body. The heat required 
to raise the temperature of a body or to increase its sensible 
heat is thus expended in increasing the velocity of the molecules 
of the body, and hence in increasing their kinetic energy. The 
heat required to effect a change of state, or to increase the 
potential energy is expended, on the other hand, in increasing 
the average distance between the molecules, and in increasing the 
volume, of the body against whatever external forces may exist. 

The gradual expansion of a body with increase of tempera- 


ture is accounted for by assuming that as the velocity of the 
molecules is increased, their average path is increased also, and 
hence their average distance apart, and hence the volume of the 

(3) Temperature. We have seen above that temperature 
refers to the condition of a body as regards its sensible heat, or 
the kinetic energy of its molecules. Two bodies are said to be 
at the same temperature when there is no tendency for heat (that 
is, molecular energy) to flow from one to the other. This con- 
dition is measured by the thermometer, an instrument too well 
known to require particular description. 

The Fahrenheit scale which is commonly used by engineers 
in the United States is graduated as follows : The temperature 
of melting ice is called 32 and that of boiling water 212. The 
interval between the two is then evenly divided into 180 parts, 
and the same divisions are extended above and below as far as 
may be desired. 

On the Centigrade scale the temperature of melting ice is 
called o and that of 'boiling water 100. The interval is then 
evenly divided into 100 parts, and the divisions extended above 
and below as may be desired. 

For transforming temperatures from one scale to the other 
we have the following equations : 

F == 9/5 C + 32 

C == 5/9 (F -- 32) 

where F and C denote respectively the temperatures on the 
Fahrenheit and Centigrade scales. 

Examples: Transform 20 C into Fah. 

Operation : F =: 9 x 20 -^ 5 + 32 = 36 + 32 = 68. 

Transform 20 Fah. to C. 

Operation: C =5 /9 (20 32) 5/9 ( 12) - -6 2/3 
or 6 2/3 below zero. 

Transform 77 Fah. into C. 

Operation : C =5/9 (77 32) = = 5/9 x 45 = = 2 5 

Transform 22 Fah. into C. 

Operation : C == 5/9 ( 22 32) == 5/9 ( 54) - - 30 

(4) Heat Unit. Care must be taken to distinguish between 
quantity of heat and temperature. The first refers to the total 
amount of heat energy present in the substance, the second to the 
kinetic energy of a molecule. A large cup of warm water and 


a small cup of hot water may both have the same quantity of 
heat, but not the same temperature. A cup of hot water and a 
barrel full of hot water may have the same temperature, but the 
quantities of heat will be very different. 

For measuring quantities of heat, use is made of a heat unit 
defined as the amount or quantity of heat required to raise one 
pound of water one degree in temperature. Inasmuch, further- 
more, as the amount thus required would vary slightly at different 
temperatures, it is necessary to fix the temperature at which the 
unit is to be defined. The temperature thus taken for the 
definition of the heat unit has sometimes been at the freezing 
point, or again at the point of maximum density of water which 
is about 39 Fah., or again at about 62, or from 62 to 
63. It is very difficult to determine the amount of heat required 
to raise water one degree at or near the freezing point, while 
between 60 and 70, or at about an average atmospheric tem- 
perature, the measurements are most readily made. For this 
reason a temperature within this range is to be preferred for 
definition of the heat unit. . 

The heat unit thus defined is often known as the British 
Thermal Unit, the name being usually abbreviated to B. T. U. 

(5) Joule's Equivalent. Since heat is but a form of energy 
it follows that it should be possible to transform heat into me- 
chanical work, and vice versa. It is for the first purpose that 
the steam engine is used, while instances of the latter transforma- 
tion are constantly before our eyes, as in the heat developed 
by the friction of a bearing, or in turning a chip from a bar of 
steel, etc. 

It therefore becomes of importance to know the ratio of 
transformation, or how much mechanical work measured in 
foot-pounds corresponds to one heat unit as above defined. This 
has been made the subject of careful experiment extending over 
the past one hundred years, and the latest and most reliable 
results seem to give for this ratio the value 778. That is, one 
B. T. U. is equivalent to 778 foot pounds of mechanical work. 
This means that when in a steam or other heat engine heat is 
transformed into mechanical work, for every B. T. U. so trans- 
formed and disappearing as heat, 778 foot pounds of work will 
be obtained ; or again when mechanical work is transformed into 
heat, for every 778 foot pounds so transformed and disappear- 
ing as work, one B. T. U. of heat will appear. 


This number or ratio, 778, is known as the mechanical 
equivalent of heat, or frequently as Joule's equivalent, though 
its value as determined by Joule was somewhat smaller than the 
value given above. 

(6) Transfer of Heat. Heat may be transferred from one 
body or place to another in three different ways : by radiation, 
by conduction, and by convection. 

By radiation we mean the transfer of heat through space in 
straight lines, as from the sun to the earth, or from a furnace fire 
to the face when the door is opened. 

By conduction we mean the transfer of heat along a body 
from one molecule to the next, as when a slice bar becomes 
warm at one end if red-hot at the other. 

By convection we mean the transfer of heat from one point 
of a liquid or gas to another by means of currents set up in the 
liquid or gas and carrying the heated molecules from one place 
to another, as in the circulation set up within a Scotch boiler. 

Two other operations are also concerned in the transfer of 
heat from one substance to another. These are emission and 

By emission we mean the giving off or transfer of heat from 
the molecules of one body to those of another. By absorption we 
mean the converse of this, the receiving of heat by the molecules 
of one body from those of another. 

The heating of water in a boiler is due, at least in part, to 
all of these processes. The fire in the furnace radiates heat to 
the crown sheet ; convection and draft currents convey the hot 
gas to the heating surface ; there is emission from the hot gases 
and absorption by the metal of the heating surfaces; there is 
conduction through the metal from the fire to the water side ; 
there is emission from the metal and absorption by the water; 
and finally there are convection currents developed in the water 
by means of which the temperature is more or less uniformly 

[3] Steam. 

Steam or the vapor of water is the substance almost univer- 
sally used as the medium through which the heat set free by 
the coal is transformed into the mechanical work of the engine 
or pump. Its properties are therefore of great importance for 
the engineer, and we may properly study briefly the more im- 
portant at this point. 



(i) Formation of Steam. Let AB in Fig. 276 be a very tall 
vessel open at the top and having an inside cross-section of one 
square inch. Let us suppose at the start that there is in the bot- 
tom of this vessel one cubic inch of water at say 60 temperature. 
On top of the water let there be a piston as shown which we 
may suppose to move without friction, and to be without weight. 
In other words we wish simply, something to separate the water 
from the air. Then on the surface of the water there will be just 
the atmospheric pressure of say 14.7 pounds per square inch. 
Now let heat be applied at the bottom of the chamber. The first 


Fig. 276. The Formation of Steam. 

result will be a transfer of heat through into the water, and a 
resultant rise in temperature of the water near the heating sur- 
face. In consequence of this the water will expand somewhat and 
thus become lighter than the other water above and farther from 
these surfaces. The heated water will thus tend to rise to the 
top and so displace the cooler water there, which will in conse- 
quence sink and thus in turn be brought into contact with the 
heating surfaces. There is set up in this way a general ascend- 
ing current of warmer water and a corresponding descending 
current of cooler water by means of which the whole mass is 
gradually raised in temperature. 

In this manner are formed the convection currents referred 


to in the preceding section, and to their formation in a steam 
boiler is due the circulation of the water, especially in those of 
the fire-tube or tank type. 

In this way then the temperature of the water will gradually 
rise until it reaches 212. The temperature then ceases to rise 
and steam begins to form, rapidly increasing the volume below r 
the piston and thus forcing it upward. The steam formed is of 
the same temperature, 212, as the water of which it is formed, 
the only changes being the increase of volume and the change 
of state. If we continue to supply heat at the bottom and pre- 
vent its escape from the sides of the tube or chamber, the water 
will thus gradually be all converted into steam, just balancing 
by its own pressure the atmospheric pressure on the top of the 
piston. The volume of steam thus formed would be 1,663 cubic 
inches, or in other words the tube would have to be 1,663 inches 
or 138.6 feet high to allow of the operation as we have supposed 
it to take place. 

Suppose now that at the beginning the piston is loaded down 
with a weight so that the total load on the water is 20 pounds 
instead of 14.7. Then let heat be supplied as before. A like 
series of changes will follow, but the water instead of beginning 
to change state from liquid to vapor at 212 must be heated to 
228 before the change begins, while the final volume will be 
only 1,244 cubic inches. If the initial pressure were 100 pounds 
then the temperature at which the change of state would begin 
would be 328, and the final volume would be only 275 cubic 
inches. Similarly for 200 pounds pressure, the figures are 382 
and 144 cubic inches. 

On the other hand, if by means of an air pump the pressure 
on the water were decreased below that of the atmosphere the 
temperature of change would become lower, and the volume 
greater. Thus if the pressure is 10 pounds the figures are 193 
and 2,385 cubic inches, and if 5 pounds they are 162 and 4,576 
cubic inches. Similarly, to boil water at 32, or the freezing 
point, it would be necessary to reduce the pressure to .089 
pounds, while the volume of vapor formed would be about 21,170 
cubic inches. 

In general we thus find that as the pressure is increased or 
decreased there is a corresponding rise or fall in the tempera- 
ture at which the formation of vapor begins, and that it re- 
mains fixed at this temperature during the process of change of 


state, and that the temperature of the vapor formed is also the 
same as that of the water from which it is formed. 

It thus appears that steam is simply the vapor of water, 
or water in the vaporous state, as defined in [2]. The 
process of steam formation as above described is known as 
boiling or ebullition, or sometimes in the general sense as vapor- 
ization. The temperature at which the change of state occurs is 
known as the boiling point or point of ebullition. 

At lower temperatures there is always a certain amount of 
slow vaporization going on, the pressure of the vapor formed 
being limited to that corresponding to the temperature of the 
water. It is by means of such slow vaporization that water is 
carried up for the formation of rain clouds from the surfaces of 
rivers, lakes and oceans, or that mud dries up in the roads, or 
clothes when hung out to dry. Thus if the temperature is 100 
the maximum vapor pressure is about one pound. Considerable 
vapor will thus be formed, which will gradually find its way up- 
ward into the air, at least so long as the air is not saturated ; i.e., 
already charged with vapor of the full pressure corresponding to 
the temperature. This is, however, a branch of the subject that 
we cannot here pursue further. 

When, however, the temperature is such that the vapor 
pressure is sufficient to just balance the entire pressure on the 
surface of the liquid, then the vapor is formed with freedom, and 
more or less from the body of the liquid, producing the agitation 
and other conditions which constitute boiling or ebullition as re- 
ferred to above. 

We have thus examined in some detail the formation of 
steam at constant pressure. Let us next examine briefly its for- 
mation at constant volume, the case which corresponds to its 
generation in a steam boiler. 

The boiler being first open to the air through the safety 
valve, the water is subject simply to atmospheric pressure. The 
heating of the water and first formation of steam proceed accord- 
ing to the process as described above. The air is thus displaced 
from the boiler and driven out through the safety valve or other 
escape. The safety valve or escape is then closed and in con- 
sequence the steam which is formed tends to increase the pres- 
sure, and as this rises the temperature of the boiling point will 
rise also. In this way as heat is added, a part of it is used in 
raising the temperature of the water and steam up to the ever- 


rising boiling point, while the remainder goes for the vaporiza- 
tion of a fresh portion of steam. In this way the volume of vapor 
remains practically constant, but the pressure and temperature 
rise together according to the regular law which relates the one 
to the other, while the increase of vapor is accommodated in 
the constant volume by the increase in density, or decrease in 
volume required per pound as the pressure rises. 

It is thus seen that the temperature and volume of steam 
are closely dependent on the pressure to which it is subjected. 

For engineering purposes pressure may be measured from 
two different starting points. The true or so-called absolute 
pressure is measured from the zero or no-pressure condition, and 
is the true or total pressure exerted by the gas or vapor in ques- 
tion. The ordinary steam gauge, however, as described in Sec. 
17 [7] does not measure the absolute pressure, but simply the 
difference between this and the pressure of the atmosphere. This 
is due to the fact that the gauge is subjected to the pressure of 
the steam on one side of the tube and to that of the atmosphere 
on the other side. It thus measures simply the difference be- 
tween the two. This is commonly called "gauge pressure." The 
absolute pressure is greater than the gauge pressure therefore, 
by the pressure of the atmosphere. This varies with the altitude 
and with other circumstances affecting the barometer. For en- 
gineering purposes it is very commonly taken at 15 pounds per 
square inch. A more correct average for the sea level is, how- 
ever, 14.7, as noted above. It results therefore that having given 
the gauge pressure we may find the absolute pressure by adding 
15 or 14.7, as we may choose, according to the degree of ac- 
curacy needed in the case in hand. 

(2) Saturated and Superheated. Steam. When steam and 
water are present in the same vessel together and there is no 
tendency for the water to change into steam or the steam into 
water except as heat is added or taken away, the water and the 
steam are said to be in thermal equilibrium. When steam is thus 
in equilibrium in contact with water, it is said to be saturated. 
Thus during the entire process of steam formation illustrated in 
Fig. 276, the steam is saturated. 

When there is no moisture or water in the liquid condition 
suspended in or mixed with the vapor, the steam is said to be 'dry. 

When there is moisture or water suspended in or mixed 
with the vapor, the steam is said to be moist or wet. In such case 


the steam or vapor part of the mixture must itself be in the 
saturated condition as defined above, so that wet steam is simply 
a mixture of saturated water vapor and liquid water. 

When steam is free from all suspended moisture, but is still 
in the saturated condition as determined by its pressure, tem- 
perature and volume, it is called dry and saturated. Thus the 
condition of the steam in the vessel of Fig. 276, just as the last 
bit of vapor is formed, is dry and saturated. During the opera- 
tion the vapor would be dry and saturated provided it contained 
no suspended moisture. Practically this is a condition difficult 
to realize. The greater or less violence of the ebullition is apt 
to carry up a certain amount of water in the shape of fine mist 
into the steam space, from which it settles back only slowly, and 
so is liable to be carried over into the engine. The steam fur- 
nished by the average boiler under good conditions contains 
usually not less than from I to 2, per cent, of moisture, while 
under poor conditions the amount may rise to 5 per cent, and 

Suppose now, referring to Fig. 276, that after the last bit of 
water has been vaporized, heat is still further applied, the pres- 
sure on the piston remaining the same. The temperature which 
during the process of vaporization has remained stationary, will 
now begin to rise, accompanied also by an increase of volume. 

If we now recall the three stages which the water has passed 
through, all at constant pressure, it appears that during the first 
stage there was a rise in temperature of the water at nearly con- 
stant volume : during the second stage (that of steam formation) 
there was an increase of volume at constant temperature : during 
the third stage, as just described, there is increase of both volume 
and temperature. 

If in this last operation instead of keeping the pressure con- 
stant the piston were held fast, thus keeping the volume con- 
stant, we should find with the addition of heat that both the tem- 
perature and the pressure would increase. 

Steam in the condition resulting from these operations is 
said to be superheated. As compared with saturated steam its 
temperature and volume are greater for the same pressure, or its 
temperature and pressure are greater for the same volume, or 
its volume is greater and pressure less at the same temperature. 
It is clear that superheated steam cannot be in contact with 
water and remain superheated. It cannot therefore be moist. 


If it were brought into contact with water it would lose its super- 
heat and become saturated, forming by the heat given up, a 
little more vapor from the water present. 

We may also put the relation between saturated and super- 
heated steam into the following form : 

Saturated steam contains only as much heat as absolutely 
necessary for its maintenance as steam at the given pressure. 
Superheated steam at the same pressure contains more heat than 

The temperature and volume per pound for saturated steam 
correspond with the pressure as in the process of vaporization, 
and are respectively the lowest temperature and smallest volume 
at which steam of the given pressure can exist. The temperature 
and volume per pound lor superheated steam are both larger 
for the same pressure than for saturated steam. 

[4] Xotal Heat in a Substance. 

(i) Total Heat of Steam. The total heat of a body in a given 
condition is the total amount of heat required to produce this 
condition, reckoning from some starting point agreed upon. For 
steam this point is usually taken as 32 Fah., or the freezing 
point of water. The total heat per pound of steam at a given 
pressure and temperature, means then the total amount of heat, 
both sensible and latent, required to produce one pound of steam 
of the given pressure and temperature, from water at 32. These 
quantities of heat are used in the solution of problems relating 
to the heat required for evaporation, boiler efficiency, gain by 
feed-water heating, etc. The value of the total heat in terms of 
the temperature is very closely given by the following approxi- 
mate equation : 

H = 1082 + .3* 
while the latent heat is similarly given by the equation : 

L = 1114 .ft. 

In these equations, H denotes the total heat, L the latent 
heat, and t the temperature. 

Instead of using these or other similar equations, the values 
are more conveniently taken from tables prepared so as to give 
the various quantities for regularly varying values of the pres- 
sure. Thus from the Table it appears that at 14.7 pounds 
pressure absolute, or at the pressure of the atmosphere, 
it requires 180.9 B. T. U. to heat the water from 32 to the boil- 


ing point 212, and then 965.7 B. T. U. to completely vaporize 
it at this point. The great excess of the latter or latent heat over 
the former or sensible heat may thus be noted. It is also seen 
that according to the table the heat required to raise the water 
from 32 to 212 is not exactly measured by the difference in 
degrees which is 180. The excess is due to the fact that the 
B. T. U. is defined for i rise from 62 to 63, while the amount 
required to make i difference at other temperatures is slight- 
ly different, increasing on the whole for higher tempera- 
tures, so that between 32 and 212 the average is slightly 
greater than from 62 to 63. It thus results that the sensible 
heat per pound of water involved in a temperature change is' 
slightly greater than the number of degrees which measures 
such change. This difference is, however, so small that for most 
engineering purposes it may be neglected if more convenient, 
and the number of heat units per pound of water may be taken 
equal to the number of degrees difference in temperature. 

(2) Total Heat of a Mixture of Steam and Water. Steam as 
actually used is usually moist ; that is, it contains a small fraction 
of water. To find how much heat is required to produce one 
pound of such a mixture of steam and water, a given fraction 
being steam and the remainder water, we have simply to re- 
member that all of the water must be raised to the temperature 
of boiling, while only the given fraction is vaporized. If there- 
fore S denotes the sensible heat and L the latent heat, while 
x is the fraction which Is steam, or the quality of the steam as 
it is called, then the heat H required to produce one pound of 
the mixture will be : 

H S -f- xL. 

The quality ,r -is usually expressed on the percentage basis. 
The following examples will illustrate the use of the steam 
tables : 

(1) Find the sensible heat, the latent heat and the total heat 
in one pound of steam at 120 pounds gauge pressure.* 

Ans. from the table, 322.1, 866.6, 1188.7. 

(2) Find the heat in i pound of feed water at a temperature 
of 110. 

Ans. no 32 = 78. 

* For present purposes, and in the following problems, it will be suffi- 
ciently accurate to take the pressure of the atmosphere as 15 pounds per 
square inch. The absolute pressure is therefore found from the gauge 
pressure by simply adding 15. 


(3) How much heat would be required to produce the steam 
in example (i) from the feed water in (2)? 

Ans. The difference between the two or 1110.7. 

Remark: These three examples show how we may find the 
heat necessary to produce a pound of steam of given pressure 
from feed water of any given temperature. 

(4) How much heat is required to make i pound of steam 
at 150 pounds gauge pressure from feed water at 130? 

Ans. 1095.5. 

(5) How much heat is required to produce i pound of moist 
steam of 92 per cent, quality at 150 pounds gauge pressure from 
feed water at 120? 

Solution : 

The sensible heat per pound of steam is. ... 338.4 
The sensible heat per pound of feed is 88.0 

The difference is 250.4 

The latent heat for one pound is 855.1. But since only 92 

per cent, is vaporized, only .92 of this will be required. We have 

therefore : 

H = 5 + xL 250.4 + .92 X 855.1 = 1037. Ans. 

(6) One engine requires per I.H.P. per hour, 16 pounds 
of steam of 96 per cent, quality at 150 pounds gauge pressure, 
the feed being at a temperature of 140. Another engine requires 
20 pounds of steam of 90 per cent, quality at no pounds gauge 
pressure, the feed being at a temperature of 110. 

Find the amounts of heat required per hour in the two en- 
gines, and hence their real comparison as heat engines. 

By the methods illustrated above we find as follows : 

For the first engine 16821 B. T. U. per hour. 

For the second engine 20436 B. T. U. per hour. 

The second engine requires therefore 3615 B. T. U. per 
hour more than the first, and in consequence is about 21.5 per 
cent, more expensive in terms of the heat required. 

Further illustrations of these principles will be found in 
Sec. 58. 

[5] I/atent Heat in Passing from Ice to Water. 

We have seen in [2] (i) that a certain amount of heat is 
absorbed in the melting of ice at the constant temperature of 32. 
It is thus rendered latent or is taken up in effecting the change 
of state in the same general manner that heat is rendered latent 


in passing from liquid to vapor. It may be of value to note here 
the quantity of heat thus rendered latent. 

This amounts to about 143 heat units, or B. T. U., per pound 
of ice melted from 32 into water at the same temperature. 


[i] General Principles. 

There may be several different bases for boiler economy ac- 
cording to the particular feature held in especial prominence. 
The output of the boiler is estimated in terms of the steam pro- 
duced, and we may have the following kinds of economy : 

(1) Economy in coal consumption, increasing with the out- 
put of steam per pound of coal burned. 

(2) Economy in weight of boiler, increasing with the output 
of steam per pound of boiler. 

(3) Economy in first cost, increasing with the output of 
steam per dollar invested in the boilers. 

(4) Economy of maintenance or total life, increasing as the 
life .of the boiler is longer and the amount necessary for repairs 
is smaller. 

It is never possible to fulfil in the highest degree the condi- 
tions for these various kinds of economy, and a compromise must 
always be made among them, though usually (i) or (2) will take 
the first place in the order of importance. 

Without special note, however, the term economy is under- 
stood to refer to (i), though the considerations relating to the 
others should always be kept in mind. In some cases, as in tor- 
pedo boat and like practice, (2) may assume first place in the 
order of importance, and, perhaps, require some sacrifice relative 
to the others. We will now consider more especially the econ- 
omy referred to under (i). 

From the standpoint of coal economy or efficiency, the 
boiler is charged with all the coal that is thrown through the 
furnace doors, and is credited with the steam which it sends to 
the engine. Or to state the matter more definitely, it is charged 
with all the heat which could be gotten from this coal by perfect 
and complete combustion, and is credited with the heat which is 
transferred through and actually used in the formation of steam. 
If the efficiency were perfect, or if there were no loss, these two 
amounts of heat would be equal. Actually there are many losses, 
large and small, and, in consequence, the latter is considerably 


less than the former. The ratio of the two is known as the boiler 
efficiency. In practice its value varies from 50 to 75 or 80 per 
cent. Following are the more important sources of loss which 
occasion this drop in efficiency. 

In,, jthe first place, a little of the fuel may fall unburned 
through the grate into the ash pit. Again, a little in the form of 
dust and small bits may be carried by a strong draft, either un- 
burnt gr only partially burnt, through into the tubes, uptakes or 
funnel. Still another small portion may escape as srnoke, which 
consists almost entirely of very fine particles of unburnt carbon 
formed from the gases which are distilled away from the coal in 
the process of combustion. Still another portion of these gases 
may escape unchanged and unconsumed. Again, there may be 
an incomplete combustion of the carbon forming carbon monoxide, 
and giving only about 4,450 heat units per pound, instead of 
14,500 which result from the complete combustion into carbon 
dioxide. Hence, whatever carbon escapes in the form of carbon 
monoxide is only partly burned, and may be considered as car- 
rying away over two-thirds of the heat which would be liberated 
by complete combustion. These losses all occur in the furnace, 
and are due to poor firing and to imperfect combustion. 

To reduce them to the lowest limit, the fireman must know 
his business, and be willing to attend to it with ceaseless care 
and diligence. In addition, there must be provided, by proper 
design, the necessary supply of air both above and below the 
grate, together with such arrangements as experience may show 
are needed for good combustion with the fuel in hand. At best 
this loss may be reduced to perhaps 2 or 3 per cent., while with 
carelessness or poor design, or both, it may easily reach values 
from 10 to 20 per cent. 

The heat being thus more or less perfectly liberated- in the 
furnace, is then passed on to the boiler heating surface, whose 
duty it is to transfer it through into the water on the other side. 
The energy is still to exist as heat, but it is to be transferred from 
the hot gas to the water, thus converting- the latter into steam. 
This, however, cannot be perfectly accomplished, and thus arises 
a further loss. A part of the heat, instead of passing through the 
heating surface, goes up the funnel carried bv the escaping gases, 
and so gets away into the outside air. Another and smaller part 
escapes by radiation into the fire room. These losses it is im- 
possible wholly to avoid. It would be necessary to avoid all loss 


of heat by radiation, and to reduce the temperature of the pro- 
ducts of combustion in the funnel to that of the outside air. The 
latter, especially, cannot be done for the best of reasons. In 
the first place, the temperature cannot be reduced below that of 
the steam and water in the boiler, because heat always flows 
naturally from a hot body to a cooler one, and it will, therefore, 
flow from the gas to the water only so long as the latter is the 
cooler of the two. The actual temperature of the escaping gases 
must be considerably higher than that of the steam, because in 
the first place sufficient heating surface to reduce them to nearly 
the same temperature could hardly be allowed ; and, again, aside 
from blowers, the strength of draft is dependent on the tempera- 
ture of the hot gas in the funnel, and for a satisfactory rate of 
combustion it is necessary to discharge the products of combus- 
tion at temperatures not less than 500 to 600 degrees. This loss 
is one therefore which exists in the nature of things, and cannot 
be reduced below some 20 or 30 per cent. 

On the whole, then, the entire losses under the best condi- 
tions can hardly be reduced much below 25 per cent., while wi h 
poor conditions they may aggregate 40 to 50 per cent. The 're- 
maining fraction, or the 50 to 75 or 80 per cent., represents then 
the efficiency of the boiler as defined above. 

Since a pound of average good coal has available sor-e 
13,000 to 14,000 heat units, it follows that the heat actually 
utilized per pound of coal is usually found between say 7,000 and 
11,000 units. 

In general the conditions favorable- to high efficiency are 
the following: 

(1) A free burning coal of good quality, with suitable fur- 
naces and air supply for complete combustion. 

(2) Moderate draft. 

(3) Abundant heating surface. 

Or, as a combination of (2) and (3), we may put : 
(2) Moderate evaporation required per square foot of heat- 
ing surface. 

The opposite of these conditions will cause necessarily a loss 
in efficiency more or less pronounced according to circumstances. 

[2] Evaporation Per Pound of Coal. 

The efficiency of a boiler is often roughly estimated by the 
number of pounds of water evaporated into steam per pound of 


coal burned on the grates. This, according to conditions, may 
vary from 6 or 7 to perhaps n. Remembering that it usually 
requires rather more than 1,000 heat units per pound of steam, 
the general agreement between these figures and those above foi 
the heat utilized per pound of coal is readily seen. In fact, the 
figures for the heat utilized are derived really from a measure- 
ment of the pounds of steam evaporated per pound of coal, 
together with a knowledge of the heat required per pound of 
steam, the latter being derived, of course, from the conditions 
of the evaporation. 

When we remember the great difference in the amount of 
heat required per pound of steam, depending on the temperature 
of the feed, the temperature of the steam, and whether the steam 
is moist or dry, it is clear that for any fair measure of boiler per- 
formance in terms of steam formed per pound of coal, these 
differences must be allowed for, especially in comparisons be- 
tween boilers -working under different conditions. 

To this end it is customary to reduce the number of pounds 
evaporated to what it would be if the steam were dry and the 
temperature of both feed and steam were 212 degrees. In such 
case it would require to make one pound of steam simply the 
latent heat at 212 degrees, or 966 B. T. U. (British thermal 

This is known as the reduced evaporation, or the equivalent 
evaporation from and at 212 degrees. It is really the number of 
pounds of steam which would be formed if each required 966 
B. T. U., and is, therefore, simply a measure of the B. T. U. put 
into the steam per pound of coal. 

The ratio between the number of B. T. U. actually required 
and the 966 is known as the factor of evaporation. These factors 
are often arranged in tabular form, assuming dry steam in each 
case, but with temperature of feed water and steam varying over 
the usual rano-e. 

The annli cation of these various principles relating to boiler 
economy will be better understood by the solution of the follow- 
ing illustrative problems. In these problems we shall still for 
convenience consider the pressure of the atmosphere as 15 
pounds per square inch, and the absolute pressure therefore as 
15 pounds greater than the gauge pressure. 

(i) Temperature of feed 110, steam pressure 150 pounds, 
gauge. Thermal value of the coal 14,000 thermal units per 


pound. Efficiency of boiler .64. Find the pounds of water evap- 
orated into dry steam per pound of coal. 

Operation : 

From the steam tables : 

Heat in the water at boiling point 3384 

Heat in the feed = 1 10 32 = 78 

Difference 260.4 

Latent heat 855.1 

Heat required per pound of dry steam IIT 5-5 

Heat available per pound of coal = .64 X 14,000 = 8,960. 

Hence pounds of steam evaporated = 8960 -f- 1115.5 
8.03 Ans. 

(2) Temperature of feed 140. Steam pressure 200 pounds, 
gauge. Quality of steam 96 per cent. Thermal value of coal 
14,400 thermal units per pound. Efficiency of boiler .70. Find 
the pounds of water evaporated per pound of coal into steam of 
the given quality. 

Operation : 

Heat in the water at boiling point 361.3 

Heat in the feed = 140 32 108 

Difference 253.3 

Latent heat per pound = 838.9. 

Take .96 of this 805.3 

Heat required per pound of moist steam , . . . 1058.6 

Heat available per pound of coal = .70 X 14,400 10,080. 

Hence pounds of steam evaporated = 10,080 ~ 1058.6 

(3) Temperature of feed 100. Steam pressure 120 pounds, 
gauge. Evaporation 8 pounds per pound coal. Assuming dry 
steam, what is the evaporation from and at 212, and what is the 
factor of evaporation? 

Operation : 

Heat in the water at boiling point 322.1 

Heat in the feed = 100 32 = 68.0 

Difference 254.1 


Latent heat per pound 866.6 

Heat required per pound of dry steam 1 120.7 

Heat utilized per pound of coal = 8 X 1120.7 8965.6. 

Evaporation from and at 212 = 8965.7 966 = 9.28. 

Factor of evaporation = 1120.7 -f- 966 = 1.16. 

Evidently also 

Equivalent evaporation 8 X 1.16 = 9.28. 

(4) Same as in example (3) but assuming the steam of 95 
per cent quality, what are the results? 

Operation : 

Heat in the water at boiling point 322.1 

Heat in the feed = 100 32 = 68.0 

Difference 254.1 

Latent heat per pound = 866.6. 

Take .95 of this 823.3 

Heat required per pound of moist steam IO774 

Factor of evaporation = 1077.4 -r- 966 = 1.115. 

Equivalent evaporation = 8 X 1.11$ = 8.92. 

(5) Temperature of feed 160. Steam pressure 200 pounds, 
gauge. Quality of steam 97 per cent. Evaporation 8.8 pounds 
steam per pound coal. What is the equivalent evaporation 
from and at 212, and what is the factor of evaporation? 

Operation : 

Heat in the water at boiling point 361.3 

Heat in the feed = 160 32 = 128.0 

Difference 233.3 

Latent heat per pound = 838.9. 

Take .97 of this 813.7 

Heat required per pound of moist steam 1047.0 

Factor of evaporation = 1047.0 -r- 966 = 1.084. 

Equivalent evaporation = 8.8 X 1.084 9-54- 

(6) Compare the economy in (3) and (5). 
Ans. in the ratio 9.28 : 9.54 or i : 1.028. 

(7) Which is the more economical of the following cases : 
(a) Coal at $4.00 per ton, 9.2 pounds of steam per pound of 


coal ; quality of steam 98 per cent. ; steam pressure 150 pounds, 
gauge; temperature of feed 110. 

(b) Coal at $3.20 per ton; 8.0 pounds of steam per pound 
of coal ; quality of steam 96 per cent. ; steam pressure 135 pounds, 
gauge ; temperature of feed 120. 

By the methods of the preceding examples we find that the 
equivalent evaporations in the two cases are as follows : 

(a) : 10.46. 

(b) : "8.85. 

The cost of steam in the two cases will therefore be in the 
compound ratio. (See Part II., Sec. 6 [2]) : 

(a) : (b) : : 4.00 : 3.20. 
: : 8.85 : 10.46. 

or (a) : (b) : : 4.00 X 8.85 : 10.46 X 3-2O. 
Whence (a) : (b) : : 1.058 : I 

or case (a) is nearly 6 per cent, more expensive than (b). 

NOTE. In solving the above examples the heat in the water 
at boiling point has been taken from the tables. Somewhat more 
quickly such problems may be solved by taking the heat in the 
water at boiling point as measured simply by the difference in 
temperatures, as explained in Sec. 57 [4] (i). The difference 
is small and is very commonly neglected. In the above examples, 
however, we have preferred to use the more exact values. The 
simpler form of solution will be illustrated by solving example 
(5) in this way, and the result will serve to show the nature of 
the difference between the two. We have thus : 

Temperature of steam 387.7 

Temperature of feed 160 

To raise one pound feed to boiling point 227.7 

Latent heat for one pound = 839.9. 

Take .97 of this 813.7 

Heat required per pound moist steam 1041.4 

Factor of evaporation 1041.4 -^ q.66 1.077. 
Equivalent evaporation = 1.078 X 8.8 = 9.49. 

[3] Evaporation Per Pound of Combustible. 

In discussing further the details of stenm boiler perform- 
ance, it is often desirable to make the necessary allowance for 
the ash in the coal, or for the ash and moisture, so as to obtain 


the evaporation per pound of actual combustible matter. To 
this end it is only necessary to divide the evaporation per pound 
of coal determined as above by the fraction of the coal which 
is combustible, or what is the same thing, we may increase the 
result per pound of coal in the ratio in which the coal is greater 
than the combustible. The result will be the evaporation per 
pound of coal exclusive of ash or of ash and moisture. This 
result may relate, of course, either to the actual evaporation 
under the given conditions, or to the equivalent from and at 212. 
This will be illustrated by the following examples : 

(8) In example (5) Suppose the ash to be 12 per cent., what 
are the actual and equivalent evaporations per pound of moist 

Ans : Actual evaporation = 8.8 -- .88 = 10 Ibs. 

Equivalent evaporation = 9.54 -i- .88 = 10.84 Ibs. 

(9) Under the same conditions suppose the ash and mois- 
ture to be 15 per cent., what are the evaporations per pound of 
dry combustible? 

Ans: Actual evaporation = 8.8 -f- .85 = 10.35. 

Equivalent evaporation = 9.54 -r- .85 = 11.22. 


[i] General Principles. 

In the following discussion it will be assumed that the 
reader has a general knowledge of the chief properties of steam 
and of its relation to the heat which it contains. We will then 
take up in an elementary way a discussion of the principles 
governing its economical use in a steam engine. 

At the very outset it must be clearly understood that we 
derive the work of the engine from the heat which the steam 
contains, and not from the steam in itself. The steam is simply 
a carrier for the heat and the operation of the engine is simply 
a means for transforming into useful work a fraction of the heat 
which comes into the engine, and then rejecting the remainder 
with the steam which is its carrier. The larger the fraction of 
the heat which can be transformed into useful work the better 
the efficiency of the engine, and the constant aim is therefore 
to turn into useful work the largest possible fraction of the heat 
which enters with the steam. 

It may be asked, why not turn all the heat into work, and 
so realize a perfect efficiency? Unfortunately a series of natural 


laws and limitations seems to prevent all hope of realizing such 
an ideal, and actually we must be content with turning into useful 
work a comparatively small fraction of the total heat supplied. 
First and foremost among the causes of this reduction in effi- 
ciency is a principle or law sometimes known as the second law 
of thermodynamics. This law fixes a limit on the fraction of 
heat which can be transformed into useful work, such limit de- 
pending on the extreme temperatures between which the sub- 
stance is worked in the engine. Thus if U is the temperature 
of the steam at admission and t* that at exhaust, so that h and t* 
are the two temperatures between which the steam is worked, 
and (/i fr) is the range, then the law asserts that no engine, 
no matter how perfect, can transform into useful work a fraction 
of the entering heat greater than (fr -- fc) -f- (h + 461). As 
another way of stating this relation, the temperature may be 
supposed to be measured from a point 461, or more accurately 
460.7 degrees, below the ordinary zero of the Fah. scale. This is 
called the absolute zero, and temperature measured from this 
zero is called absolute temperature. The difference of the tem- 
peratures would be the same no matter whether measured from 
the ordinary of absolute zero. The numerator of the above frac- 
tion is therefore the difference or range of temperature, while the 
denominator is the absolute temperature of the entering steam. 
The fraction of heat converted into useful work can therefore 
never exceed the temperature range divided by the absolute tem- 
perature of the entering steam. Thus to illustrate, suppose that h 
= 370 and ts = 140. Then the fraction becomes (370 140) -4- 
(370 -f- 461 or 230 -i- 831 = .277 or slightly over one-quarter. 

These figures represent the limits for steam of about 160 
pounds gauge pressure, and it therefore appears that for engines 
operating between these limits this law steps in, and at one 
stroke reduces the ideal efficiency from one to about one-quar- 
ter ; or in other words we are forced, due to the operation of this 
law, to throw away about three-quarters of the total heat, and 
at the very best with the most ideally perfect engine could only 
transform into useful work the remaining one-quarter. 

Such then is the very best that could be done by a so-called 
ideal engine. The working substance in the simplest form of 
such an engine must be carried through a series of changes or 
operations, four in number, and specified as follows : 

(i) The first operation must consist of an expansion at con- 


stant temperature, and all heat received from the source of sup- 
ply must be received during this operation. 

(2) The second operation must consist of an expansion 
with decrease of temperature during which, however, no heat 
as such is allowed to either enter or leave the substance. 

(3) The third operation must consist of a compression 
during which the temperature remains constant and all heat re- 
moved from the body must be removed during this operation. 

(4) The fourth operation must consist of a compression with 
increase of temperature, during which, however, no heat as such 
is allowed to either enter or leave the substance, and at the end 
of which the substance must find itself in the same condition 
as at the beginning of number (i). 

Work is done by the substance during operations (i) and 
(2) and work must be done on the substance during (3) and (4). 
The difference between the work done by and on the substance 
will be the net work obtained from the heat in the substance, 
and the ratio of this to the total heat supplied during number (i) 
or the efficiency of the engine will be exactly measured by the 
difference between the temperatures of operations (i) and (3) di- 
vided by the larger increased by 461 ; or in symbols : 

; 'i *. 

efficiency = - - - 

t t + 461 

This is then the cycle and the efficiency of an ideal engine 
in the simplest form. There may be certain related variations in 
the operations (2) and (4) making a more complicated cycle, but 
with the same efficiency. This ideal marks, then, the highest 
possible limit of efficiency for any and all engines working be- 
tween the given temperature limits ft and ft. 

In the table are shown the values of this limiting efficiency 
for engines with gauge pressure as indicated, all condensing and 
supposed to have a back pressure of 2.8 pounds absolute, or a 
lower temperature, ft, of 140. 

An examination of this table shows that with the ideal con- 
ditions which correspond to the operation of this engine, the 
fraction of heat utilized with modern boiler pressures would 
range from 25 to 30 per cent. These conditions, however, are 
far from those which actually exist in practice. Every one of 
the conditions specified above is violated in greater or less 
degree, and the result is that with the operation of the engine 
under the best conditions obtainable in actual practice, the frac- 


tion realized will be only some 60 to 80 per cent, of the figures 
for the ideal case as given in the table below. These figures, 60 
to 80 per cent, in the best practice, really measure the efficiency 
of the engine so far as the engineer is responsible. That is, 
nothing which he can do will serve to avoid the loss which re- 
duces the limiting efficiency down to that for the ideal engine 
as given in the table above. His efforts are therefore limited 
to approaching as nearly as possible to the conditions of the 
ideal engine, and the figures 60 to 80 per cent, measure the de- 
gree of approach which modern engineering practice has made 


Gauge Pressure 
at Engine. 






1 2O 









2 7 8 



















to this ideal. Thus, for illustration, if the ideal engine could 
transform 25 per cent, of the heat into useful work, a good actual 
engine working between the same temperature limits will be 
able to transform from 15 to 20 per cent., and similarly for other 

To put the matter a little differently, any and all engines 
fail to transform into work all of the heat supplied to them. In 
the ideal engine as specified above, the part not transformed but 
rejected as heat is the least possible for all engines working be- 
tween the same limits of temperature /< and h. In any actual 
engine the amount not transformed into work but rejected as 


heat is greater than in the ideal case. Such additional amounts 
of heat rejected and not transformed into work are called wastes 
or losses. That is ; all differences between the performances of 
the ideal and actual engines are considered to be due to these so- 
called wastes or losses. 

These various wastes may be classified as follows : 

(a) Radiation and Conduction Waste. 

This consists of heat which is radiated away from the hot 
surfaces of the cylinder, or conducted away through the columns 
and bed plate. The heat thus escaping avoids transformation 
into work and is therefore counted as a heat waste, or as an ex- 
pense from which no corresponding return is received. 

(b) Initial Condensation. 

At the instant the steam valve opens, the steam rushes into 
the cylinder to find itself in contact with surfaces which have but 
recently been exposed to the influence of the condenser or ex- 
ternal air. They are, therefore, at a temperature much lower 
than the steam, and in consequence a part of the heat is absorbed 
and a corresponding part of the steam is condensed. The heat 
thus absorbed by the surfaces of the cylinder and piston will be 
given up later during the exhaust period of the revolution, and 
thus communicated to the condenser. It thus appears that a 
thin skin of metal on the inside of the cylinder and on the faces 
of the cylinder head and piston may be considered in a sense 
as a place of hiding into which a portion of the heat slips on the 
entrance of the steam, and from which it escapes to the con- 
denser or air without having taken part in the cycle of the en- 
gine, and hence without having contributed its part to the usefu) 
work done. The heat so escaping appears thus as an expense, 
but without any corresponding return in work, and therefore con- 
stitutes a heat waste. 

(c) Irregularities of the Cycle. 

We have specified above the four fundamental operations of 
the ideal engine cycle in its simplest form. In the actual engine 
none of these is realized, and the variations are all such as to 
count against the efficiency. Into the details of these points we 
cannot here enter, and the broad statement must suffice that 
with but few exceptions, the variations from the routine specified 
above for the ideal engine will count against the efficiency and 
occasion a heat waste greater or smaller as the circumstances 
mav determine. 


The Improvement of the Steam Engine. From the preceding 
section it follows that there are two fundamental methods open 
for the improvement of the steam engine from the standpoint of 

(1) An increase in the temperature range and thus an in- 
crease in the ideal or limiting efficiency. 

(2) The saving of some of the various wastes as noted 

The first raises the ideal efficiency and hence with a given 
proportion of wastes will raise the actual efficiency as well. 
The second raises the actual efficiency by carrying it a little 
nearer to the ideal. 

The temperature range may be increased in two ways, the 
initial temperature can be raised, and< the final temperature can 
be lowerd. 

The continually advancing pressures in modern practice 
means a constant rise in the upper temperature, a constant in- 
crease in the ideal efficiency, and with the same proportion of 
losses, a corresponding rise in the actual efficiency. This is then 
the real significance of high pressures in modern practice so far 
as they are related to the question of economy. 

Again by decreasing the back pressure from say 18 pounds 
for a non-condensing engine to say 3 pounds for a condensing 
engine, a very considerable decrease in the final temperature is 
obtained, a corresponding increase in the temperature range, and 
a resultant increase in actual efficiency. This is likewise the 
real significance of the influence of the condenser on the econ- 
omy of the engine. 

In general then, the proportion of heat wastes being the 
same, the economy will be better as the initial pressure is higher, 
and the back pressure is lower; or in general, as the range of 
pressure and temperature worked through is the greater. 

We may turn next to the problem of reducing the wastes of 
the actual engine, as specified under the three heads above. 

The waste due to radiation and conduction cannot be wholly 
avoided, but the former, which is by far the larger of the two, 
may be much reduced by suitable lagging or non-conducting 
covering. With such provision the loss under this head is 
usually very small compared with the other losses mentioned. 

The waste due to the so-called initial condensation is one 
which may be reduced, but not wholly avoided. Before discuss- 


ing the means suitable to this end some further explanation of 
the nature of the loss will be required. 

As already pointed out, the action of the metal walls in pro- 
ducing this loss depends on their capacity when at a lower tem- 
perature, for absorbing heat from the steam (as during admis- 
sion) and for giving it up when at a higher temperature (as dur- 
ing exhaust). The action depends, then, on the range of tem- 
perature between admission and exhaust, and on the particular 
readiness with which the walls absorb and reject heat according 
as they are cooler or hotter than their surroundings. There are 
therefore two distinct features to be considered the range of 
temperature, and the readiness with which the iron absorbs and 
rejects heat under the conditions mentioned. 

Now it is found that if the expansion through the entire 
temperature range is split up into a series of steps, each carried 
out in a cylinder by itself, the loss under consideration is less 
than if the entire expansion should take place in one cylinder. 
Carrying out this principle we have, of course, the multiple ex- 
pansion engine with its total range of operation divided among 
several cylinders in series. 

This, then, is the real significance of the multiple expansion 
(compound, triple, quadruple, etc.) engine, so far as its relation 
to economy is concerned the splitting of the total expansion or 
of the total temperature range into a series of steps is, found to 
reduce considerably one of the wastes, and so raise the actual 
efficiency of the engine. 

Next turning to the other controlling feature of this loss, 
the readiness of absorption and emission it seems to be the 
case that once the internal surfaces become wetted or covered 
with a film of moisture, the absorption and emission of heat into 
and from the metal proceed with much greater readiness than 
when they are dry. In other words the passage of heat between 
metal with a moistened surface and moist steam is much more 
rapid than between the same metal with a dry surface and dry or 
superheated steam. 

In the ordinary steam engine we have, therefore, an action 
of the walls due to the range of temperature employed, and to 
the natural capacity of cast-iron or steel to absorb and emit 
heat from and to the steam. This is further greatly aug- 
mented by the presence of a more or less complete film 
or layer of water over the surface, which arises from 


the condensation of the first entering saturated steam. 

The use of superheaters, reheaters and jackets is found in a 
general way to decrease the readiness with which heat ex- 
changes occur between the metal and the steam, and thus to 
decrease the amount of waste due to their actions. Thus in an 
engine using moderately superheated steam we should have the 
same general tendencies as noted above for the operation with 
saturated steam, but less augmented because of the smaller 
amount of moisture formed. In an engine using steam super- 
heated to such an extent as to remain above the point of satura- 
tion during its entire passage through the cylinders, no moisture 
is formed and the action of the surfaces is limited to that which 
can take place between their dry surfaces and the dry super- 
heated steam. The office of superheating is then simply to re- 
duce the readiness with which the exchange of the heat between 
the metal surfaces and the steam is effected. The results show 
that in such case the reduction is real and productive of a con- 
siderable increase in economy. 

Regarding the use of reheaters it seems likely that their bene- 
ficial action will be well marked in proportion as they are able to 
superheat the steam passing through them, and thus act as a 
superheater in stages, each for the cylinder next beyond. 

The beneficial results gained by the use of steam jackets are 
in large measure due to action of the same character. The 
jacket containing steam at a temperature higher than that in the 
cylinder transfers heat into the inner surface of the cylinder walls 
and thus tends to keep it dry and to reduce the amount of heat 
exchange, and hence the corresponding waste. The steam 
jacket acts also to some extent to modify the character of the 
cycle as noted below, but most of its useful action may probably 
be put down to the hindering of heat exchanges between the 
walls and the steam in the cylinder. 

It must not be forgotten, however, that whatever gain is 
thus effected within the cylinder is obtained at the expense of 
the heat drawn from the jacket, and the whole operation is there- 
fore an attempt to reduce one loss by introducing another. If 
the latter is less than the saving in the cylinder, the net result will 
be a gain equal to their difference. If the latter is the greater of 
the two, the net result will be a loss, and if they are equal the net 
result will be no change in the economy of the engine. These re- 
lations account for the varying experience with jackets, but it 


now seems well assured that when properly fitted and operated, 
the result will show a gain of from 5 to 10 per cent, over similar 
conditions unjacketed. 

We come now to the last principal division of the wastes of 
the actual steam engine, those due to irregularities in the cycle, 
or in other words to variations from the routine of operations 
which would give the efficiency of the ideal engine as discussed 
above. In this respect but little can be done to improve matters. 
The use of jackets and reheaters may possibly affect the routine 
in such a way as to bring it somewhat closer to the ideal condi- 
tions, but this is by no means certain, and the benefit due to 
these appliances comes mostly from the decrease in cylinder con- 
densation as explained above. 

There are methods, however, of modifying the cycle of 
the engine by the use of a series feed water heater, in such way 
as to bring it somewhat nearer to the ideal cycle. Such a feed 
heater for a quadruple expansion engine may consist of say three 
chambers or heaters through which the feed passes in series. In 
the first it is heated by steam drawn from the L.P. receiver. It 
then passes on to the second chamber, where it is heated by 
steam drawn from the second I. P. receiver, and then goes on to 
the third chamber, where it meets with steam drawn from the 
first I. P. receiver. As the feed water thus becomes hotter and 
hotter it meets with steam of higher and higher temperature 
drawn from the successive higher receivers in the engine, and it 
is thus brought nearly to the temperature of the water in the 
boiler. The exhaust from pumps may also be turned into the 
first chamber, thus making it a means of taking heat from their 
exhaust and of returning it with the feed to the boiler. In some 
cases also live steam of full boiler pressure has been used in a 
last chamber to still further raise the temperature of the feed. 
Various modifications may be worked out in the details of the 
operation of such feed heaters, but in all cases their significance 
lies in the fact that the cycle of operations as a whole may in this 
way be brought a step nearer to the ideal cycle than would other- 
wise be the case. All such changes, if made in accordance with 
the proper principles, may therefore result in a saving of heat and 
in a gain in the economy of the engine, and in this fact lies the 
chief significance of the feed-water heater as a feature of modern 
engineering practice. See also Section 30 with reference to the 
same subject. 



[2] Relation of Expansion to Economy. 

The question of the expansion of steam and its influence on 
engine economy has long held an important place as perhaps the 
chief factor in engine economy, and it may therefore be well to 
refer to this feature in somewhat further detail. From the stand- 
point of the preceding discussion of the question we should say 
that the expansion of steam is favorable to economy because it 
brings the cycle of operations nearer to that for the ideal engine. 
These points may, however, be treated differently and in a more 
elementary manner, and we therefore proceed to discuss the 
question by the actual comparison of the two indicator diagrams 
given by an engine working with and without expansion. 

Consider the two cards C D I H and C D E F H of Fig. 277. 
The first is the card that would be given by using the steam in 
an engine with stroke C D following to the extreme end, and 
then exhausting along D I H. The second card is such as would 



Fig. 277. The Expansion of Steam. 


be given by the same steam used expansively cut-off taking 
place at D and the stroke continuing expansively to E. The 
back pressure in each case is represented by the line H F. The 
area C D I H represents the work in one case and the area C D 
E F H in the other, so that the difference or D E F I represents 
the gain by expansive working. In other words, if we use the 
steam full pressure up to D and then exhaust it, we throw away 
an amount of work measured by D E F I which might be saved 
by expansive working. It is likewise true that the exhaust open- 
ing at E causes a loss of work E J F which might be saved by 
continuing the expansion down until the forward pressure falls 
to the back pressure as at J. It is rare, however, that we can 
afford cylinders of sufficient size to carry the expansion to sucfi 
a point, and, as may be seen, the amount thus lost is smaller 


and smaller as the final pressure E G is nearer the back pres- 
sure F G. 

To illustrate the gain by expansive working let us suppose 
the initial pressure L C = 100 pounds, the back pressure L H 
= 3 pounds, and that the cut-off is at a series of points .1 .2 .3, 
etc. Then neglecting the effect due to clearance, the number of 
expansions will be as 10, 5, 3.3, etc., and the ratio of saving 
will be as given in the following table. The numbers in the 
column headed e give values of the ratio DEFI-i-CDIH, 
or the ratio of the amount saved by expansion to the amount 
done before expansion begins. 

Point of Cut-off. Expansion Ratio. e. 

.1 10. 2. II 

2 5- 1-55 

3 3-30 I.I8 

4 2.50 .91 

.5 2.00 .69 

.6 1.66 .51 

7 143 -36 

.8 1.25 .23 

.9 i. ii .11 

.10 i.oo .00 

It will be understood that the figures of the above table refer 
to indicator cards such as those of the diagram in which there 
is no allowance for clearance, compression, rounding off of cor- 
ners, etc. These conditions are of course taken in order to 
simplify the necessary computations. The nature of the results 
would, however, be the same in the actual case, and these figures 
may therefore be taken as a sufficiently close indication for il- 
lustrative purposes. 

It thus appears that the gain is proportionately greater the 
larger the number of expansions, and for the highest efficiency 
we should therefore carry the expansion to the highest limit. 
Practically there are two considerations which fix an early limit 
to this extension of the expansion range. The first is the limit 
of size. The greater the number of expansions the larger the 
engine and hence, especially for marine engines, we can seldom 
afford weight enough to give the number of expansions which 
other considerations might warrant. The second limitation 
comes from the increase of internal or cylinder condensation 


which increases with the number of expansions until finally the 
resulting waste would off-set the gain due to the increase in 
ideal efficiency. 

This loss is decreased by the compounding of engines, so 
called, or by the splitting up of the total expansion into a series 
of steps in separate cylinders. Hence with multiple expansion 
engines we are able to employ higher rates of expansion without 
corresponding losses from cylinder condensation than with a 
single cylinder; and this, as we have seen in [i], is the real sig- 
nificance of the use of multiple expansion rather than simple 

[3] Economy of the Actual Engine. 

We have already seen that the highest possible efficiency of 
an ideal engine under usual conditions will be found between 25 
and 30 per cent., while the actual engine at the best will realize 
only some 60 or 70 per cent, of these figures, or an efficiency of 
say 15 to 20 per cent, in good practice. Now one horse power 
is 33,006 foot-pounds of work per minute, and from the value of 
the work equivalent of heat this is equal to 33,000 -r- 778 = 
42.42 heat units per minute. Hence one horse power means the 
transformation of 42.42 heat units per minute into mechanical 
work. It follows that the heat which must be supplied to the en- 
gine in order to provide for one horse power will be given by di- 
viding the number 42.42 by the efficiency at which the transforma- 
tion is effected. But 42.42 -f- 15 = 283. + and 42.42 -f- 20 = 
212. -J-. Hence, placing the limits a little more broadly, it ap- 
pears that in good practice we shall require from say 200 to 300 
heat units per minute for each horse power developed in the en- 
gine. This corresponds to a range of 12,000 to 18,000 heat units 
per hour. Now remembering that each pound of steam brings to 
the engine roughly 1,000 heat units, it is clear that this will cor- 
respond to a range of steam consumption of say 12 to 18 pounds 
per I. H. P. per hour. These figures may be taken as covering 
the range of good practice from about the best at present attain- 
able to a value only moderately fair for modern triple expan- 
sion engines, or good for the usual type of compounds. 

Again each pound of coal burned may be expected to fur- 
nish under good conditions some 9,000 or 10,000 heat units to 
the water in the boiler, or to transform some 9 or 10 pounds of 
water into steam. Hence the coal required per I. H. P. per 
hour will be given by dividing the heat units or the pounds of 


steam required by the amount of either which may be expected 
from one pound of coal. This will give a coal consumption of 
from about 1.2 to 1.8 pounds per I. H. P. per hour, which may 
also be considered as representing the upper part of the range 
of good practice for triple and quadruple expansion engines 
under from moderately good to the best conditions at present 
attainable. In a few exceptional cases by the use of feed heaters, 
superheated steam and all means favorable to economy the con- 
sumption has been reduced to i.o pound coal per I. H. P. per 

For compound and simple condensing engines under good 
to moderate conditions the steam consumption will rise to from 
20 to 30 pounds of steam, corresponding roughly to from 2 to 3 
pounds of coal with good boiler economy, and to perhaps 2.5 to 
3.5 pounds with poor boiler economy. 

Farther along the line will come engines perhaps non-con- 
densing ; and of still lower efficiency, such for example as electric 
light, centrifugal pump, blower, winch, and steering engines. 
The steam required for them may rise to from 40 to 60 pounds or 
more per I. H. P. per hour, corresponding to a coal consump- 
tion of from perhaps 4 to 8 pounds, according to the boiler 

Still lower in the scale of economy we find the ordinary 
direct acting pump. Such pumps operate in the steam cylinder 
with almost no steam expansion, and the piston speed is very 
low, thus giving full time for the transfers of heat which cause 
cylinder condensation. Due to these and other less important 
causes the steam consumption may rise to 200 pounds and more 
per I. H. P. per hour, while rarely can it be brought as low as 
100 pounds. This corresponds to a coal consumption from say 
10 to 25 pounds, depending somewhat on the efficiency of the 
boiler. In terms of absolute efficiency these figures correspond 
to from about I to 3 per cent., the values thus ranging down- 
ward from the 15 to 20 per cent, given above as the highest 
values at present attainable. 


As we have already seen in Sec. 59 the coal required per 
I. H. P. per hour in good practice is usually found between say 
1.5 and 2.0 pounds. Where especial attention is given to econ- 


omy the figure may be reduced below the lower value down even 
to i pound per I. H. P. per hour, while by the neglect of due 
attention, or in cases where the conditions are such that econ- 
omy must be sacrificed, the value may rise above the higher 

Let : c denote the pounds of coal per I. H. P. per hour. 
H denote the I. H. P. 

Then cH = pounds of coal per hour, 

cH -i- 2,240 = tons of coal per hour, 

, 24 cH 3 cH cH 

and - or - - or - - = tons of coal per day. 

2240 280 93.3 

As a thumb rule for a quick estimate, we may remember 
that at a coal consumption of 1.86 pounds per I. H. P. per hour 
(a figure only moderately good), the coal required per day will 
be 20 tons per 1,000 I. H. P. 

The use of the above formulae may be illustrated by the 
following examples : 

(i) With a coal consumption of 1.78, how much coal will 
be required in the bunkers of a ship making a seven-day trip, the 
I. H. P. being 2,400 and a margin of 10 per cent, being allowed 
for emergencies? 

^ * 3x1.78x2400 

Coal per day = - - = 45.75 tons. 


Coal for 7 days = 7 X 45-75 = 320.25, say 320 tons. 

Margin 32 tons. 

Coal in bunkers = 352 tons. 

(2) Which will require the more coal per day, a ship with 
9,800 I. H. P. at 1.82 Ib. per I. H. P. per hour, or two ships each 
of 4,000 I. H. P. at 2.20 Ib. per I. H. P. per hour? 
For the first : 

3X 1.82X9800 

For the second : 

3X 2.20x8000 

= 191 tons per day. 
= 188.5 tons per day. 


Difference 2.5 tons. 

(3) How long time can a vessel steam on 213 tons of coal 
and how far on a speed of 12 knots, the I. H. P. being 3,600 and 
the coal consumption being 1.68? 



Coal per hour = - -==2.7 tons. 


Time 213 -4- 2.7 = 78.9 hours. 
Distance = 78.9 X 12 = 947 miles. 

(4) A vessel's log shows 420 tons of coal used in a period of 
9 days, 16 hours. The average I. H. P. was 2,120. What was 
the coal consumption per I. H. P. per hour? 

Number of hours 9 X 24 -\- 16 = 232. 

Amount used per hour = 4055 Ib. 


Coal consumption = ^5- = I.QI Ib. 


As a further development of the same problem we may often 
wish to find the coal burned per mile, or per ton-mile of displace- 
ment, or per ton-mile of cargo. These we may illustrate by the 
following examples : 

(5) Given: Displacement... = 9,486 tons. 

I. H. P = 12,000 

Speed = 18 knots. 

Coal consump- 
tion = i .8 Ib. per I. H. P. per hour. 

Cargo = 2,000 tons. 


Coal per hour = 1.8 X 12,000 = 21,600 Ib. 

Coal per mile = 21,600 -f- 18 = 1,200 Ib. 

Coal per ton-mile of disp. = 1,200 -4- 9,486 = .127 Ib. 

Coal per ton-mile of cargo = 1,200 -4- 2,000 = .6 Ib. 

(6) If the same ship were to be driven at but half the speed, 
only about one-eighth the I. H. P. would be required, or say 
1,500 I. H. P., while the cargo might be increased to say 5,000 

With the same engine economy we should then require : 
Coal per hour 1.8 X 1,500 = 2,700 Ib. 
Coal per mile = 2,700 -f- 9 = 300 Ib. 

Coal per ton-mile of displacement 300 -4- 9,486 = 
.316 Ib. 

Coal per ton-mile of cargo = 300 -f- 5,000 = .06 Ib. 

(7) Again a case similar to one of the large modern ocean 
freighters : 

Displacement = 27,000 tons. 
Speed = 13 knots. 


I. H. P. = 6,600. 

Cargo = 15,000 tons. 

Coal consumption =1.3 lb. per I. H. P. per hour. 


Coal per hour = 1.3 X 6,600 = 8,580 lb. 

Coal per mile = 8,580 -i- 13 = 660 lb. 

Coal per ton-mile of disp. = 660 -f- 27,000 = .0244 lb. 

Coal per ton-mile of cargo = 660 -4- 15,000 = .044 lb. 

At the other extreme take a torpedo-boat of the destroyer 
type as follows : 

Displacement = 310 tons. 

I. H. P. = 6,200. 

Speed = 31 knots. 

Coal consumption = 2.2 lb. per I. H. P. per hour. 

Then : 

Coal per hour = 2.2 X 6,200 = 13,640 lb. 

Coal per mile = 13,640 -f- 31 = 440 lb. 

Coal per ton-mile of disp. = 440 -j- 310 = 1.42 lb. 

These examples illustrate the principle that per ton-mile, 
less coal is burned as the ship is larger and goes slower, while 
more is burned as she is smaller and goes faster. This is the re- 
sult of the two facts. 

(1) As the ship increases in size the power required per 
ton of displacement for a given speed decreases, and accordingly 
the larger 1 the ship the less the coal required per ton at a given 

(2) For a given ship, as the speed increases, the power and 
hence the coal required increase nearly as the cube of the speed 
ratio, while the time for a mile or for a given voyage is reduced 
only in the simple ratio of the speeds. Thus to illustrate : If 
the speed is increased 10 per cent., or say from 10 knots to n 
knots, then the power will be increased nearly in the ratio 
(n -i- io) 3 =3 (i.i) 3 = 1.331, while the time on the mile or on a 
given voyage will be decreased in the ratio io -i- u. Hence the 

coal will be changed in the compound ratio -- x --- or 1.21 

to i. Hence it appears that in such case the increase of speed in 
the ratio i.i to i will increase the coal per mile or per voyage in 
the ratio 1.21 to i. Or briefly an increase of io per cent, in the 
speed will mean an increase of about 20 per cent, in the total 
coal required. Similarly, of course, a decrease of io per cent in 


the speed would mean a decrease of about 20 per cent, in the 
total coal required.- 

If there were no other principle involved, it would follow 
that the slower a given vessel goes the more cheaply could she 
make a given voyage. This would be true if we could consider 
only the power in the main engine, and assume for it a constant 
coal economy. This cannot be done, however, in the case 
of a single ship going at different speeds, because as the power 
in the main engines is decreased below its normal amount the 
coal required per I. H. P. increases continuously. Furthermore 
the power required for the various auxiliaries never decreases 
in the same ratio as the power of the main engines, and for cer- 
tain auxiliaries the power is hardly affected by the change in 
the main engine. The coal required for the auxiliaries becomes 
therefore greater and greater relative to that required for the 
main engine. Due to these facts it follows that as the speed is 
decreased a point will be reached below which the saving in the 
total coal required per hour will be more than offset by the in- 
crease in the time required, so that for a given voyage the total 
coal expense will begin to increase rather than decrease. This 
point is known as the "most economical speed" and is the speed 
at which a given voyage can be made with the least expenditure 
of coal. Its value will depend largely on the amount and char- 
acter of the auxiliary machinery in operation, but is often found 
at a speed somewhat above half the full power speed. In the 
mercantile marine it is rare that ships are operated at speeds 
other than those corresponding to normal full power conditions, 
so that the determination of a most economical speed is not of 
great importance in such cases. In the naval service, however, 
where economy may be of more importance than the reduction 
of time required for a voyage, ships are often operated at or 
about the most economical speed, and its determination and the 
principles fixing its location are of importance in such cases. 


The spring loaded safety valve as described in Sec. 17 [i] is 
used almost exclusively in modern practice. The ability to solve 
problems relating to the weighted arm safety valve (see also 
the same section), is, however, required of all candidates for 
U. S. Engineer's Licenses, so that it is of importance to 


thoroughly understand the method of solving the various prob- 
lems which may arise in this connection. These problems are 
all special cases of the general problem in mechanics which has 
to deal with the equilibrium of a body under the action of a 
system of forces, and for a clear understanding of the matter 
the principles discussed and explained in Part II., Sec 12, must 
be kept well in mind. 

The arrangement of a weighted arm safety valve is shown 
in skeleton in Fig. 278. The steam presses upward on the lower 
face of the valve V, and is opposed by three forces tending to 
keep the valve on its seat as follows : 

(1) The weight of valve and spindle direct. 

(2) The weight of the lever with center of gravity at some 
point N and pivoted at the fulcrum O. 

(3) The weight proper at M acting with a leverage or arm 
equal to MO. 

Now just as the valve is about to open, these two sets of 



1 t 




Fig. 278. The Safety Valve Problem. 

forces, the one acting upward and the other acting downward, 
will balance. It is furthermore a principle of mechanics that 
when such forces are just on a balance, the product of the forces 
by their arms or leverages must make the same sum in each 
direction. Thus in the present case the up force at S may be 
considered as tending to cause motion of the lever about the 
fulcrum O, in the direction of the hands of a watch, while the 
down forces at B and W tend to cause motion about the same 
fulcrum in the opposite direction. We therefore measure the 
arms from the fulcrum point O. 

Now if A is the area of the valve in square inches and /> the 
steam pressure in pounds per square inch by the gauge, the 
total steam load on the valve will be the product pA. This acts 
directly upward, and, as noted above, is directly opposed, as far 
as it goes, by the weight of the valve and spindle. Denote this 
weight by V. Then the difference (pA-V) is the actual or net 
force transmitted from the valve to the lever at S, and tending, 
as noted above, to turn the lever about O from left to right. Let 


a denote the arm for this force, or the distance SO from the 
center of the spindle to the center of the fulcrum. Then (pA-V)a 
is the product of force by arm for the upward forces. Let us 
save this and turn next to the remaining or downward forces. 
Let W denote the amount of weight at M, and / the arm or dis- 
tance MO from the center of gravity of the weight to the iui- 
crum. Also let B denote the weight of the lever, and c the arm 
or distance NO, from the center of gravity of the lever to the 
fulcrum. Then (Wl + Be) is the sum of the products for the 
downward forces. By condition these are equal when the two 
sets balance and the valve may be considered as on the point 
of opening. Hence as an equation we shall have : 

Wl + Be = (pA-y)a or 

Wl + Bc= p'Aa-Va. 

From this equation we can find the value of any one quan j 
tity which we may wish, provided we know all the others. Thus 
suppose we know all but W. Then we have : 

l[7 _pAa - VaBc ( , 

7 ............. V 1 ; 

Similarly if we know all but / we have : 
pAa- Va Bc 


and if we know all but f( the pressure per square inch at which 
the valve will open with a given weight and location, we have : 

We may readily express by rules the operations represented 
by these equations as follows : 

Rule (i) To find the weight knowing the other quantities. 
Multiply together the pressure per square inch, the area of the 
valve in square inches, and the distance from the center of the 
valve spindle to the center of the fulcrum. From this subtract 
the product of the weight of the valve and spindle by the same 
arm SO, and also the product of the weight of the lever by its 
arm NO. Divide the remainder by the weight arm MO, and the 
quotient will be the weight desired. 

Rule (2) To find the location of the weight or length of the 
arm MO knowing the other quantities. 

Find the same difference as in rule (i) and divide by the 
weight W ' . The quotient will be the length of the arm MO. 


Rule (3) To find the pressure at which a given valve and 
weight will lift. 

Multiply the weight W by its arm MO ; also the weight of 
the lever by its arm NO, and the weight of the valve and spindle 
by its arm SO. Add these three products and divide the sum by 
the product of the area of the valve times the arm SO. The 
quotient will be the pressure desired. 

Example : 

Let MO or / = 28 in. 

Let NO or c = 12 in. 

Let SO or a = 4 in. 

Let diameter of valve = 3^ in. 

Then area of valve or A = 9.62 sq. in. 

Let weight of lever or B = 5^2 lb. 

Let weight of valve or V = 4 lb. 

Let steam pressure or p 80 per lb. per sq. in. 

Then, pAa = 80 X 9.62 X 4 = 3078.4. 

Va = 4 X 4 = 16. 

Bc=$y 2 X 12 = 66. 

Then 3078.4 16-66 = 2996.4 and W = 2996.4 -f- 28 = 
107 lb. in round numbers. 

Or if W were known and / required, we should find the same 
numerator 2996.4, and then have : 

/ = 2996.4 -7- 107 = 28 inches in round numbers. 
Or if p is desired we should have : 

Wl = 107 X 28 = 2996 

Be = $y 2 X 12 = 66 

Va = 4X4 = 16 


Aa = 9.62 X 4 = 38.48- 

We then have : 

p =3078 -r- 38.48 = 80 lb. in round numbers. 

General Remarks on the Problems. The arm / is to be meas- 
ured from the center of gravity of the weight W to the fulcrum 
or turning point O. Usually the weight is of regular form, cir- 
cular or rectangular in elevation, so that its center ot gravity is 
readily found. If the lever turns about a pin, then the arm / 
must be measured to the center of the pin. If it is provided with 


a link and knife edge bearing as in Fig. 78 then / is measured to 
the bearing edge. If the center of gravity of the weight W 
and the fulcrum O are not on the same horizontal line, then the 
arm / must be measured as the horizontal distance between ver- 
ticals drawn through these points. 

The center of gravity of the lever arm must be obtained 
practically either by measurement or by balancing on an edge 
in the familiar manner. If it is practically a uniform straight bar 
the method of measurement will be quite accurate ; if it is taper- 
ing or irregular the method by balancing may be preferred. In 
any event, with usual proportions, as seen in the example above, 
the influence of the lever is relatively small, so that a slight error 
in the values relating to its weight or center of gravity would be 
of much less importance than a like proportional error in the 
weight W or its arm /. 

The area of the valve, A, should be that, of course, of the 
lower face, or more accurately, of the opening at the lower edge 
of the seat. 

The weights of the various parts are of course obtained by 
weighing. If this is not practicable, a fair approximation may 
be made by computation based on careful measurement. In such 
case the volumes are found by the most appropriate means ac- 
cording to the shape of the figure (see Sec. 77) and then by the 
use of the known weights of the substances per unit volume, 
(see p. 30) the weights may be. found. 

The general U. S. regulations relating to safety valves will 
be found among the extracts from the Rules of the U. S. Board of 
Supervising Inspectors given in Sec. 19. 


As we have seen in Sec. 16 all flat surfaces of any consider- 
able size, in a boiler, require some support in addition to that 
which can be furnished by their own strength. In fact the whole 
idea of bracing is to subdivide by the braces a large flat sur- 
face into a sufficient number of smaller surfaces, each of which 
shall be self-supporting between the points where the braces are 
connected to the plate. The braces are then designed so as to 
be able to carry the entire load as a whole, and the parts of the 
plate between the braces are simply required to support, without 
undue change of form, the part of the load which comes upon 



To illustrate by a diagram let Fig. 279 represent a part of a 
boiler head requiring bracing. Now imagine the plate entirely 
cut out around the dotted line, and then fitted in so exactly as to 
make a steam tight joint. The part thus cut out is therefore 
entirely separated from the remainder of the head, and without 
some especial support would be blown out immediately when 
steam was raised. Now suppose the braces to be so designed 
that they may be safely depended on to carry the entire load on 
the plate, and thus keep it securely in place in the head. It 
simply remains then for the parts of the plate between the braces 
to support themselves without losing their proper shape, and 
the support is thus made complete. In the actual boiler head, or 
in all cases where a flat surface has to resist pressure, the design 
of the braces is worked out exactly along these lines, and no 

/ Aooooooooo 
/ / ooooooooo 
I /oooooooooo 
1 / oooooooooo 






ooooooooo \ 
oooooooooo \ 
oooooooooo \ 


Fig. 279. Boiler Bracing. 

account is taken of the strength of the plate for the general sup- 
port of the load as a whole. 

In designing boiler braces we have to consider two things : 

(1) The total load to be supported and number of braces, 
or simply the load upon one brace as determined by their spac- 
ing and the steam pressure. 

(2) The safe load per square inch of section of brace. 

The total load depends on the area to be supported and on 
the gauge pressure. In figuring out the area, as in Fig. 279, it 
is customary to consider that a narrow strip of metal around the 
outside will be well supported by the shell or by the tubes. The 
width of such strip is usually taken as 2 or 3 inches, though 
there seems to be no good reason why it should not be taken 
as half the spacing or pitch of the braces. This amounts to con- 



sidering the shell and tubes as effective bracing or support lor 
that part of the plate near them, in the same manner and to the 
same extent as for the braces themselves. 

If the area thus found is multiplied by the gauge pressure, 
the total load results. The spacing of the braces must then be 
taken in accordance with the principles and rules given in Sec. 19. 
The total number of braces is thus determined, and the average 
load per brace may be found by dividing the total load by the 
number. Or otherwise, after the spacing is decided upon, the 
load on each brace may be found by multiplying the area which 
it supports by the pressure per square inch. Thus in Fig. 28oa 
the surface supported by the brace is considered as the dotted 
square, and the spacing being the same both ways, its area 
equals the square of the pitch. In some cases the spacing is not 
the same in both directions, as in Fig. 28ob. In such case the 

Marint Engineering 

Fig. 280. Boiler Bracing. 

supported area is found by multiplying together the two pitches. 
Sometimes, again, the braces are irregularly distributed, and the 
area supported by one brace may be roughly triangular or of 
other irregular shape. In such case the area which the brace may 
be fairly called upon to support must be taken by approxima- 
tion, using 1 the best judgment which can be brought to bear on 
the special circumstances. 

When the braces are arranged in rows and columns as in 
Fig. 279, it will usually be found that due to the necessary spac- 
ing about the edges, the average load is slightly less than that 
which would correspond to an entire square or rectangle, and in 
consequence it is usually safer to take the load as determined 
directly by the area supported. Thus in Fig. 28oa, let the side 
of the square be 7 inches : then the supported area is 49 square 
inches. Likewise in Fig. 28ob, let the pitch be 14 inches in one 


direction and 16 inches in the other: then the supported area 
equals 14 X 16 or 224 square inches. Again in Fig. 67, suppose 
that three braces are to be used to support the approximately 
triangular area on the back tube sheet. In such case we make 
a fair allowance for the support about the edges, sketch in the 
area which the braces may be called on to support, sketch in the 
braces so as to divide the area as evenly as possible, and either 
compute the area of the whole triangle and divide by 3, or com- 
pute the smaller areas separately. 

Turning now to the second chief question, the safe load to 
be allowed per square inch of section of brace, we find that the 
U. S. Rules provide that iron braces shall not be allowed more 
than 6,000 Ib. per square inch of section ; while for steel, if in- 
spected according to regulation, the allowance may be as fol- 
lows : From \y$ in. diameter to 2^2 in. diameter, not to exceed 
8,000 Ib.; and above 2^/2 in. diameter, not to exceed 9,000 Ib., 
each per square inch of section. (See also Sec. 19). 

It must be noted that in all cases the diameter is measured at 
the root of the thread or at the smallest section. For this reason 
the threads are usually raised so that the diameter at the bottom 
is not less than that of the body of the brace. 

We have thus discussed the determination of two necessary 
items the load which the brace is to support, and the safe load 
per square inch of section. It is clear then that if we divide the 
latter into the former, the quotient will be the necessary cross 
sectional area of brace. Having found the area, we find the 
diameter by means of a table of diameters and areas, or by the 
proper rule or formula of mensuration. See Part II., Sec. 9 [10]. 

These various operations may be expressed in the form of a 
rule as follows : 

(1) Find the area to be supported by one brace, and multiply 
this by the gauge pressure per square inch. The product will 
be the load to be supported by the brace. 

(2) Take the safe load per square inch of section in accord- 
ance with the rule above given. 

(3) Divide the total load as found in (i) by the safe load as 
taken in (2) and the quotient will be the necessary area in 
square inches. 

(4) Find the corresponding diameter either by the help of a 
suitable table or by means of the proper formula or rule of men- 


To illustrate the foregoing a few examples will be of aid. 

(1) In Fig. 279, suppose the total area to be supported is 
found by measurement to be 3,784 square inches, the steam 
pressure being 160 pounds gauge. Find the total load. 

Ans. Load = 3784 x 160 = 605,440. 

(2) In Fig. 28oa, suppose the braces spaced 14 inches each 
way, the steam pressure being the same as in (i). Determine 
the load on one brace. 

Ans. Load =: 14 x 14 x 160 = 31,360. 

(3) Suppose, instead, that we wished to space the braces 14 
inches one way and 16 inches the other. Find the load on one 

Ans. Load = 14 x i6x 160 = 35,840. 

(4) Suppose that we space screw staybolts 6 inches x 6 
inches. Find the load on one brace. 

Ans. Load =6 x6x 160 5,760. 

(5) Suppose in (4) that the spacing were 7x6^. Find the 

Ans. Load = 7 x 6 l /> x 160 = 7,280. 

(6) What would be the area and diameter of a steel brace in 
(2), allowing 8,000 Ib. per square inch of section? 

Ans. Area = 31,360 -r- 8,000 = 3.92 square inches. 
Corresponding diameter = 2% inches nearly. 

(7) What would be the area and diameter of a steel brace in 
(3), allowing 8,000 Ib. per square inch of section? 

Ans. Area = 35,840 -i- 8,000 = 4.48 square inches. 
Corresponding diameter = 2 7-16 scant. Probably 2^ 
inches would be employed. 

(8) What would be the area and diameter of an iron stay in 
(4), allowing 6,000 Ib. per square inch of section? 

Ans. Area = 5,760 -h 6,000 = .96 square inches. 
Corresponding diameter = i 1-8 inches nearly. 

(9) What would be the area and diameter of an iron stay in 
(5), allowing 6,000 Ib. per square inch of section? 

Ans. Area = 7,280 -~ 6,000 = 1.213. 
Corresponding diameter = i*4 inches nearly. 

(10) Suppose that screw stay bolts are spaced 7x7 inches, 
the steam pressure being 200 Ib. gauge. Find the area and diam- 
eter of a steel stay, allowing 8,000 Ib. per square inch of section. 

Load = 7 x 7 x 200 = 9,800 Ib. 

Area = 9,800 -H 8,000 = 1.225 square inches. 



Corresponding diameter = i j4 inches nearly. 

Load on Oblique Braces. In case the brace is not at right 
angles to the surface to be supported, proper allowance must be 
made for the increase of load on the brace due to the angle of 
obliquity. This problem is explained in Part II., Sec. 12 
[14] (14)- 

Load on Forked. Ends of a Brace. The load on the forked ends 
of a brace, as in Fig. 59, is greater than one-half the load on the 
brace, in a ratio depending on the angle of obliquity. This prob- 
lem is explained in Part II., Section 12 [14] (13). 


In order to examine the relation of the strength of a boiler 
shell to its diameter, thickness and the steam pressure, consider 

3 F H D 

Uttriiie t^'iyiiuxrinff 

Fig. 281. The Stnt gth of Rollers. 

first a hollow chamber, as in Fig. 281, with parallel sides, AC and 
BD, a face AB perpendicular to these sides, and for the other end 
any other curved or irregular surface CD. Let this contain 
steam under pressure. Now it is a well-known fact of experience 
that under such circumstances the chamber will remain in equi- 
librium, and it will not move as a whole, and in particular will 
move neither to the right nor to the left. Hence the internal 
force acting to the right must equal that to the left. But the 
force acting to the right is the total resultant of all the forces act- 
ing on the curved surface CD, while the force acting to the left 
is the resultant of the parallel forces acting on the plane face AB. 
Hence numerically these two resultant* must be equal, nnd this 
will be the same, no matter what the shape of the surface on the 
right, as, for example, EF or GH. Now AB is called the pro- 
jected area of any curved surface, such as CD, EF or GH, the 


direction of projection being of course parallel to the sides AC 
and BD. Hence we may say that the total resultant force in any 
direction due to the pressure of steam or of any gas or vapor act- 
ing on the curved surface will equal the pressure of the projected 
area of such surface, the projection being taken in the direction 
of the resultant desired. This is a very general and very im- 
portant principle in mechanics, and has many applications, one of 
which is to the problem of the strength of a boiler, as we will pro- 
ceed to show. 

In Fig. 282 let ABCD denote a cross section of a cylindrical 
boiler with steam pressure acting on the curved surface, as de- 
noted by the arrows. Suppose a plane of division AB, and let us 
consider what it is which keeps the two halves from separating 
under the action of the steam pressure. The surface ACB is 

Q Marine Engineering 

Fig. 282. The Strength of Boilers. 

urged upward and the surface ADB is urged downward, while 
they are prevented from separating by the strength of the ma- 
terial at A and B. 

Now the force tending to thus separate the two parts is evi- 
dently measured by the force urging ACB upward or ADB down- 
ward. As we have just seen, this equals the force on the pro- 
jected area which is represented by AB. Suppose the axial 
length of the section which we are considering to be one unit, or 
one inch, and denote the diameter by d, the radius by r and the 
thickness of the shell by t. Then the area of AB equals 'd. square 
inches, and if p is the pressure per unit area the total load on AB 
is pd. But as we have seen this is numerically the same as the 


force which urges ACB upward and ADB downward, and 
which is opposed simply by the strength of the material at A arid 
B. The cross sectional area on each side will be t x i or / ; hence 
the total area of material will be 2t. Let 5 denote the stress de- 
veloped in the material per square inch of section. Then 2/5* is 
the total stress developed, and this must equal the load pd* 
Hence we have the equation 2tS = pd = 2/>r. (i) 

2/5 tS 

and p = ~~d = 7 ^ 3) 

In these equations (3) gives the value of the steam pressure 
p, which would produce the stress S in the metal of thickness t. 
If, however, a shell, as in Fig. 282, were formed with riveted 
joints the strength of the metal in the joint would be less than 
that of the plate itself in the ratio given by the efficiency of the 
joint as discussed in section 15. Hence, if 5 is to be the safe 
working stress in the metal of the joint, and e is the efficiency, 
the working pressure p must be reduced in the ratio of the effi- 
ciency, or from p to ep, in order to keep the stress in the joint 
down to the value S. Also if T is the ultimate strength of the 
metal, we do not wish S to rise above a certain fraction of T. 
The number by which we divide T to find the safe stress S is 
called the factor of safety. Denote this factor by f. Then in (3) 
substituting for 5" its value T -f- f and allowing for the efficiency 
of the joint we have : 

2etT etT 
P = ~j j r = - f (4) 

and , m (5) 

In a similar manner let us consider the strength of the boiler 
to withstand rupture around the shell. In this case the area of the 
head is nd 2 H- 4 and the load is/nv/ 2 -~ 4. The section of metal 
carrying this load is measured by the circumference multiplied by 
the thickness, or by xdt. Hence if 5 is the stress developed per 
unit area, the total stress in the metal carrying the above load is 
ndts. Equating the load and the total developed stress we have : 

7T dtS = 


4/5 2/5 

or P ~ 


Comparing this with (3) it is seen that in the case of the 
shell without seam or joint the pressure necessary to produce 
rupture around the circumference will be just twice that required 
for rupture along the length ; or, in other words, the boiler is 
twice as strong for rupture around the circumference as for rup- 
ture along the length, and this is an important principle which 
should be borne in mind in dealing with questions relating to the 
strength of a cylinder against pressure from within. It also fol- 
lows that the longitudinal seams must be made with the greatest 
care and of the highest efficiency, while joints of lower efficiency, 
so long as they insure tightness against leagage of steam, will 
be sufficient for the circumferential seams. To take account of 
the factor of safety and of the efficiency of the joint we must in- 
troduce in (6) the factors / and' e, the same as in (3). This 
will give : 

4 etT 2etT 
P = ~fd = 7~^ (?) 

and , = (8) 

For a bumped boiler head, as referred to in section 19, we 
consider that the head is a part of a sphere, and that all parts of 
such a surface are equally strong. Now for a sphere as a whole 
we have for the total load on a circumferential section the pres- 
sure p multiplied by the projected area of the hemisphere. But 
the latter is ^ 2 ~ 4, and therefore the load is ^pd^ -f- 4. The 
total section of metal is ndt, snd if 5 is the stress developed per 
unit area, then the total stress is ndtS. Hence we have : 


= 7T dtS 



or p = 

But, as seen above, this is the same as the value for a cylin- 
drical shell for rupture around the circumference. Hence we 
have the principle that a sphere has the same strength in all di- 
rections as a cylinder of equal diameter for rupture around the 
circumference. It will be noted that this relates simply to the 
strength of a sphere or part of a sphere for pressure on the con- 
cave surface. For the strength of a head bumped inward or con- 
vex on the inside, there is no method of treatment by simple 

For the mechanics involved in the computations relating to 


plain boiler bracing reference may be made to section 62 and to 
Part II., section 12. 


In the days of the jet condenser, and when blowing off to 
reduce the density of the water in the boilers was the usual prac- 
tice, the loss of heat occasioned by this operation was necessarily 
the subject of consideration, and it became necessary to be able 
to compute this loss in any given case. This is most easily done 
by the simple application of algebraic methods. 

Let F denote the pounds of feed water in any given time, 
and f its density. 

Let B denote the pounds of water blown out in the same 
time and b its density. 

Then (F 5) = pounds of water evaporated into steam in the 
same time. Likewise fF represents the amount of solid matter 
brought into the boiler during the giveif time, and bB represents 
similarly the amount blown out in the same time. Since the 
density of the water in the boiler remains constant at b, the 
amount of solid matter in the water must remain constant, and 
hence as much must be blown out as comes in by the feed, or : 
fF = bB (i) 

From this we readily derive the following relations : 
F b 
B =? ^ 

^B~ = "/" (3) 
Now let ti = temperature of feed, 

/a = temperature of steam, 

H = total heat in one pound of steam at given 

Then B (/-/) = heat blown out, and (F-B) [H-(t^2)] = 
heat put into the steam formed. 

Then (F B) (H + 32) + Bt* Fti = total heat. 
Hence ratio of loss e is given by : 

e = X (', f,} ^ 

By the aid of the ratios above, this expression is readily re- 
duced to the following form : 

^? __ 


These algebraic operations may be expressed by the fol- 
lowing : 

Rule (i) Multiply the density of the feed water by the dif- 
ference between the temperatures of the steam and of the feed. 

(2) Subtract the density of the water blown out from the 
density of the feed. 

(3) Add 32 to the total heat of i Ib. steam. 

(4) Multiply together the results in (2) and (3). 

(5) Multiply the density of the feed by the temperature of 
the steam. 

(6) Multiply the density of the water blown out by the tem- 
perature of the feed. 

(7) Add the results in (4) an<I (5) and subtract from the sum 
the result in (6). 

(8) Divide the result in (i) by that in (7) and the quotient 
expressed in per cent will give the percentge of loss. 

Examples : (i) Density of feed, f=i. 

Density maintained in boiler, 6 = 2. 
Pressure of steam = 100 pounds, gauge. 
Temperature of feed ti = 100. 
Then from tables : t* = 337.8, 
andH= 1185. 
337.8 100 

Then loss ratio, e = 

(1185 + 32) 4-337.8 2 X ioo 

or e= = 17.5 per cent. 

We may follow through the details somewhat differently, as 
follows : 

The loss of heat per pound of water blown off equals (it fr). 
This equals 337.8 ioo or 237.8. 

The heat required per pound of water evaporated is 
H (U 32). This equals 1185 (100-32) or 1117. Now from (3) 
it appears that the amount evaporated is to the amount blown 
out as (b /) is to f or as i is to i. That is, the amount evaporated 
equals the amount blown out. Hence for every 1117 heat units 
put into a pound of steam 237.8 are lost. Hence the percentage 
loss on the total heat employed is 237.8 -f- (237.8 + 1117) 

or e = -^' = 17.5 per cent, as before. 

o D i * 

(2) Density of feed, f = ft. 

Density maintained in boiler, b = ift. 


Steam pressure = 60 pounds gauge. 
Temperature of feed ft = 92. 
Then from tables : ft = 307.4, 
and H = 1175.7. 
Then percentage of loss, 
I (307-4 92) 

c ~ 

f (ii75-7 + 3 2 ) +iX 307-4 if X 92 


or f - - ?- - = - = jS.A per 

6x1207.74-7x307.4 13X92 8202 


Or again by analysis : The loss of heat per pound of water 
blown off equals (ft ft). This equals 307.4 92 = 215.4. The 
heat required per pound of water evaporated is H (ft 32). 
This equals 1175.7 (9 2 3 2 ) = 1115.7. Now from (3) it ap- 
pears that the amount evaporated is to the amount blown out as 
^4 is to % or as 6 : 7. Hence for every 6 Ibs. evaporated there 
will be 7 Ibs. blown out. The corresponding loss of heat is 
7 X 215.4 = 1507.8. The corresponding amount of heat put into 
steam is 6 X 1115. 7 = 6694.2. The total heat used is 1507.8 + 
6694.2 = 8202. The percentage of loss will be then 



18.4 per cent, as before. 


As we have seen in Sec. 18 a certain fraction of the heat is 
lost by way of the funnel. In certain forms of feed-water heat- 
ers, a part of this loss is prevented by placing the heater at the 
base of the funnel to absorb the heat of the gases after they 
have left the tubes. In water tube boilers such arrangements 
are especially common, the heater consisting usually of a con- 
tinuous coil of pipe jointed up with elbows or return bends, and 
through which the feed- water passes before going to the upper 
drum, or point of regular feed entrance. 

It thus becomes a question of interest as to how much sav- 
ing may be effected by the feed-water heater thus arranged to 
utilize a part of the heat in the waste funnel gases. This will 
be best illustrated by an example. 

(i) Temperature of feed 110. Pressure of steam 160 Ib. 
gauge. Assuming dry steam, what will be the percentage gain 
by heating the feed-water to 170? 


From Table I for the first condition : 
Total heat in i Ib. steam = .......................... 1 195 

Heat in feed-water = 1 10 32 = ---- .................. 78 

Heat required to form i Ib. steam = .................. 1117 

For the second condition the heat units saved are measured 
by the difference in the feed-water/' temperatures, or in this case 
by 170 no = 60. 

Hence 60 heat units have been saved out of 1117, and in 
this case the heat required per pound of steam formed will be 
1117 60= 1057. 

The percentage saving is measured by 60 -f- 1117 = 5.7 per 

In case the feed-water is heated by exhaust steam which is 
not. sent to the condenser, and of which the heat would be other- 
wise wasted, the gain is found in the same manner by dividing 
the rise in the temperature of the feed-water by the number of 
heat units needed to form one pound of steam without the heater. 

In case the feed-water is heated by live steam from certain 
of the receivers, or by any steam which might otherwise have 
been used or the heat of which might have been saved, then the 
question of heater economy becomes much more complicated 
and cannot be determined by any process of simple computation. 
It becomes simply a question of where it is most advantageous to 
use the heat, whether in the heater or elsewhere, a question 
which in general can only be answered by the actual trial. See 
also Sec.. 30. 


The total expansion of the steam in the multiple expansion 
engine is attained by expanding it in the high pressure cylinder 
from the point of cut-off to the end of the stroke, and then 
handing it over to a series of cylinders of continually increasing 
size until the steam which first filled the H. P. cylinder to the 
point of cut-off, finally fills the L. P. cylinder, and the expansion 
is complete. It would seem at first that the total number of 
expansions would be given by dividing the volume of the L. P. 
cylinder by that of the H. P. up to the point of cut-off. It is not 
quite true, however, that the volume of the entering steam is 
measured by the volume of the H. P. up to the point of cut-off, 


nor that its final volume is that of the L. P. cylinder. These 
simple relations are modified by the clearance in the manner 
described in Sec. 68. Due to this effect the actual number of 
expansions will usually be from .5 to i less than the apparent 
number given by dividing the H. P. volume up to the cut-off into 
the L. P. volume. 

The number of expansions suitable in any given case will 
vary with the initial steam pressure and with the other conditions 
to be fulfilled. 

With steam having an initial pressure of 150 to 180 pounds 
and used in triple expansion engines the number will usually 
vary from say 8 to 12 ; toward the lower values as the importance 
of the development of power per ton of machinery is greater, 
and the importance of coal economy is less, and toward the 
higher limit and perhaps even beyond in the reverse cases. With 
higher steam pressure, say from 180 to 220 pounds, and quadruple 
expansion engines, the number of expansions will be commonly 
found between 10 and 15, varying in one direction or the other 
according to the same general considerations as given above for 
the lower pressures. 

Of this total expansion range not more than 1.4 to 1.6 is 
usually obtained in the H. P. cylinder with the usual cut-off 
between .55 and .75 of the stroke, and taking into account the 
effect of the clearance. This leaves the remainder to be obtained 
from the ratio between the volumes of the H. P. and L. P. cylin- 
ders, and assuming the same stroke this will equal the ratio be- 
tween the areas of the cylinders. Hence with from 8 to 12 total 
expansions the ratio between the piston areas of the L. P. and 
H. P. will usually be found say from 5 to 7, while witfi a higher 
steam pressure and from 10 to 15 total expansions the ratio will 
be from say 7 to 10. 

We shall not here enter into the details of the proportions of 
the cylinders of multiple expansion engines, and it will be suffi- 
cient to add to the foregoing the following simple rules by 
which suitable values for the diameters of intermediate cylinders 
may be found having given those of the high and low. 

(a) For triple expansion engines. 

(1) Take the square root of the H. P. diameter. 

(2) Take the square loot of the L. P. diameter. 

(3) Multiply together the results of (i) and (2), and the 
result will give a value for the intermediate diameter. 


Example : Diam. of L. P. = 50". 

Diam. of H. P. = 2o".25. 

1/5Q 7.07. 
1/2025 = 4.5. 

4-5 X 7-07 = 31-8. 

It is usually considered better to take the actual value 
slightly under rather than over the value given by the rules, and 
we may therefore take 31 or 31 J^ as a suitable diameter for the 
intermediate cylinder. 

(b) For quadruple expansion engines. 

(1) Take the cube root of the H. P. diameter. 

(2) Take the cube root of the L. P. diameter. 

(3) Square the result found in (i). 

(4) Multiply together the results found in (2) and (3) and 
the product will give a value for the diameter of the first M. P. 

(5) Square the result found in (2). 

(6) Multiply together the results found in (i) and (5) and 
the product will give a value for the diameter of the second M. P. 

Example : Diam. of H. P. = 27. 
Diam. of L. P. = 80. 

1/27 = 3. 

(3) 2 = 9- 

9 X 4-31 = 38.79- 

(4.31)' = 18.58. 

3 X 18.58 = 5574. 

Here also it is usually considered better to take the actual 
diameters slightly under rather than over the values given by 
the rule. Hence in taking shop dimensions we may go under 
rather thani over, and in the present case take say 38 and 55 as 
suitable values for the diameters of the two M. P. cylinders. 


The term clearance is used in two senses. Clearance proper 
denotes the actual distance between the face of the piston and 
that of the cylinder head when the former is at the end of the 
stroke. That is, it is the least distance between the piston and 
the cylinder head. In amount it may vary from *4 to J^ or J4 
inch, being naturally larger the larger the engine. 

The clearance volume or the percentage clearance on the 
other hand is the actual volume contained between the face of the 


valve and the face of the piston when the latter is at the end of 
the stroke, or it is such volume expressed as a percentage of the 
volume swept by the piston. The clearance should be deter- 
mined either by measurement and computation from the draw- 
ings, or by filling it with water and measuring the amount re- 
quired. There are several methods of procedure. In the first 
place the valve must be disconnected and blocked in mid position, 
thus covering the ports. Care must also be taken to provide by 
the use of putty, if necessary, against leakage at either the valve 
or piston. Then place the engine on the center and by means 
of the indicator pipe fill the clearance volume with water by pail- 
fuls, weighing each pailful before pouring in, and the amount left 
over in the last pailful. Then knowing the weight of the pail, the 
total weight of water poured in may be found. This reduced to 
volume by taking 62.5 Ib. to the cubic foot will give the clearance 
volume in cubic feet, and this divided by the volume of the piston 
displacement will give the clearance percentage. If salt water 
were used, 64 instead of 62.5 would be used in reducing to 

Somewhat differently the mode of procedure may be as fol- 
lows: Place the engine just one inch off the center as shown by 
measurements on the guides. Fill up> the volume as before and 
note the weight required. Then move the engine up to the center 
slowly, catching the water as it is forced out and weighing as be- 
fore. The amount forced out corresponds to i inch of piston dis- 
placement. Subtract this amount from the total, and the re- 
mainder represents the water in the clearance. Divide the latter 
by the amount representing one inch of piston travel, and the 
quotient is the number of inches corresponding to the clearance. 
This divided by the stroke will give the clearance percentage. 

As an illustration of the first mode of procedure, suppose 
diameter = 22 inches, stroke = 40 inches, weight of water to fill 
clearance = 85 pounds. The volume of clearance = 85 -f- 62.5 
= 1.36 cubic feet. The volume of piston displacement = 3.1416 
X ii X ii X 40 -^ 1728 = 8.8 cubic feet, nearly. Hence clear- 
ance percentage 1.36 -f- 8.8 = 15.45 per cent. 

For the second mode of procedure let the figures be as fol- 
lows : Total weight of water with engine i inch off center = 99 
pounds. Weight of water forced out when engine is brought to 
center = 13.5 pounds. Difference = 85.5 pounds. Then 13.5 
pounds represents i inch of piston travel, and 85.5 pounds the 

49 6 


whole clearance. Hence 85.5 -f- 13.5 = 6.33 inches = number 
of inches of piston travel giving a volume equal to that in the 
clearance. Hence 6.33 -f- 40 inches = 15.8 per cent. = clearance 




As we have seen in Sec. 67, the clearance volume is defined 
as the volume or space between the piston when at the end oT 
the stroke and the face of the valve. It comprises the "clearance 
proper" or space between the piston when at the end of the 
stroke and the cylinder head, together with the volume of the 
ports or passages leading from the valve face to the cylinder. 

O A Marine *<,VnHn0 p 

Fig. 283. The Effect of Clearance on the Expansion Ratio. 

The volume of the clearance expressed as a fraction of the vol- 
ume swept by the piston is usually known as the clearance ratio 
or per cent., and is usually found in marine practice from .10 to 
.15, though in some cases it may rise as high as .20. The steam 
within this volume takes part, of course, in all expansions and 
compressions to which the steam in the cylinder as a whole is 
subjected, and its influence on the apparent expansion ratio must 
therefore be considered. 

If there were no clearance volume, then the expansion ratio 
would be given by dividing the total volume swept by the piston, 
by the volume up to the point of cut-off. But this would be the 
same as taking the reciprocal of the cut-off ratio. Thus, for 
example, if the cut-off were at 1/2 stroke the expansion ratio- 
would be 2 ; if at 1/3 stroke, 3; if, at 2/3 stroke, 3/2 or 1.5, etc. 
With a clearance volume, however, this is modified as shown by 
Fig. 283. Let AB denote the volume swept by the piston and 
OA the clearance volume to the same scale, or otherwise let AB 


denote the length of the stroke and OA the clearance volume re- 
duced to stroke by dividing the volume by the piston area. Then 
if cut-off is at some point X the actual volume of steam within 
the cylinder and ready to expand is denoted by OX rather than 
by AX. Again at the end of the stroke when the piston reaches 
B, the final volume of the steam is OB. Hence the real expan- 
sion ratio is OB/ OX and denoting its value by r we have : 

= AX+O A 

Now dividing both numerator and denominator of this frac- 
tion by AB we have : 

i + OA 


AB + AB 

Now AX -f- AB is the cut-off ratio, and OA -f- AB is the 
clearance ratio or per cent. Denote the first of these by a and the 
second by c. Then we have : 

i + c 

~- ^m- 

Examples : 

(1) Cut-off at */2 stroke, clearance 10 per cent. 
Find the true expansion ratio. 

Operation : a = 3/2 = .50. 

c = 10 per cent. = .10. 

i. oo + .10 1. 10 

Hence r - - ~ =1.83 Ans. 

.50 + .10 .60 

(2) Cut-off at 60 per cent, clearance 15 per cent. Find the 
true expansion ratio. 

I. oo -f . i Z i. ic 

Operation: r = i 53 Ans. 

.60 + .15 .75 


As seen in Sec. 55 [3] we have for the horse power formula : 

33,000 33,ooo 

Now the factors 2.LA -=- 33,000 are always the same for any 
one cylinder, while the other two (pN) will vary according to the 
conditions. We may therefore compute in advance the value of 
the factor (2.LA -f- 33,000) and then to find the H.P. we shall 


have simply to multiply these by the other two, of which one, p, 
is found from the cards, and the other, N, from the counter. This 
factor 2.LA -f- 33,000 is called the "engine constant," and is often 
thus computed as a matter of convenience, especially when large 
numbers of cards are to be worked up. 

For the power in one end of the cylinder only we have sim- 
ply to take the factors LA -r- 33,000 with N and the value of p 
found from the corresponding card. To allow for the area of the 
piston-rod on the lower side of the piston in the formula for the 
full power, we may use the average area top and bottom with the 
average mean effective pressure. When there is a difference in 
the values of the mean effective pressure top and bottom, this 
will not give quite the same result as if the two ends were taken 
separately. The difference, however, is in all ordinary cases quite 
unimportant. See also Sec. 55 [3] . 

Example : Given a cylinder of diam. = 36", stroke = 42", 
diam. of piston-rod = 5"- Find the constant neglecting the pis- 
ton-rod, and also allowing for it as above explained. 

Area of cylinder = 1017.9 sq. in. 

Stroke = 3.5 feet. 

Constant = (2 X 3-5 X 1017.9) -v- 33,000 = .2159. 

Next to allow for the piston-rod we have for its area 19.6 
square inches. Taking this from 1017.9 we have 998.3 square 
inches as the area of the lower side of the piston. The mean of 
the upper and lower sides is then 1008.1. It may be noted that in 
all such cases the mean may be most easily obtained by taking 
from the upper area one-half the piston-rod area, or in this case, 
by taking 9,8 from 1017.9, giving 1008.1 as above. We then 

Constant = (2 X 3-5 X 1008.1) -=- 33>ooo = .2138. 


The indicated thrust in pounds may be defined as the indi- 
cated power in foot pounds divided by the product of the pitch of 
the propeller multiplied by the revolutions per minute. 

Let: ;.j 

H = I. H. P. 

p = pitch of propeller. 

N = revolutions per minute. 

T indicated thrust. 


T in pounds = 

Then by formula : 

33,000 // 


This may be reduced to tons by dividing by 2,240, and we 

T- 33)0oo H 14.73/7 
have T in tons = - ^ , 

Dividing the above value of T in pounds by the value of the 
reduced mean effective pressure as given by equation (i) in Sec. 
71, and we have : 

T 33,000 H iLNA 2 LA 


An p 33>ooo p 

It thus appears that the ratio of the indicated thrust to the 
reduced mean effective pressure is measured by twice the stroke 
times the L.P. area divided by the pitch of the propeller. All 
of these are constants for any given engine, and it thus follows 
that the ratio between the indicated thrust and the reduced mean 
effective pressure is a constant, or in other words that the former 
is in a constant ratio to the latter. The indicated thrust which is 
often considered as a rather vague quantity is thus related to the 
reduced mean effective pressure, a much better known quantity. 

The indicated thrust may also be considered as the actual 
thrust which would be exerted if the propeller worked without 
slip, and if all the power developed in the cylinders were used in 
driving the ship forward at the speed thus produced. Actually 
a part of the power is lost in the friction of the engine and in the 
water due to the operation of the propeller, while the latter does 
not operate without slip. In consequence the actual thrust ex- 
erted on the thrust block is usually found somewhere about two- 
thirds the indicated thrust, computed as above. 


The I. H. P. is 1,640, the pitch 13 feet, and revolutions 148. 
Find the indicated thrust in pounds and in tons : 

, . 33,000 X 1.640 

7 in pounds = - - 28,130 

13 X 148 

T in tons = 28,130 -f- 2,240 = 12.56. 


In the multiple expansion engine the power, as we know, is 
developed in the various cylinders, as equally as the designer is 
able to bring about. The reduced mean effective pressure may be 
defined as the mean effective pressure which, acting; in the low 


pressure cylinder alone with the same piston speed, would pro- 

duce the same power as the actual engine with its series of cyl- 


, . Taking the usual formula for power, as in Sec. 55 [3] we 

have : 


and solving for p we have 
33,000 H 3^ 

* 2 LrtN '- (2 LA/) A 

Hence if the entire power were to be developed in the L. P. 
cylinder the necessary mean effective pressure would be found 
by the operations indicated by this equation, and such would be 
the reduced mean/ effective pressure, or the mean effective pres- 
sure reduced to the L.P. cylinder. The operations indicated by 
the above equation may be expressed by a rule as follows : 


(1) Multiply the indicated horse power by 33,000. 

(2) Multiply twice the length of the stroke in feet by the 
revolutions (giving piston speed) and this by the area of the low 
pressure piston in square inches. 

(3) Divide the result found in (i) by that found in (2) and 
the quotient is the reduced mean effective pressure desired. 

To obtain a somewhat different expression for the reduced 
mean effective pressure denote the areas of the three pistons 
H.P., I.P. and L.P. of a triple expansion engine, for example, by 
Ai, As, A*, and the corresponding values of the mean effective 
pressure in these cylinders by {H, p*, p*. Then the total power H 
of the formula (i) above may be expressed as follows : 

H _. (2LN)frA* + (*LN)pJ 9 +(2LN)fJ 3 


According to (i) this value of H is to be multiplied by 33,000 
and divided by 2 LN times A* the L.P. piston area. This will 
give the following as the value of the reduced mean effective 
pressure : 

According to this formula the procedure for finding the re- 
duced mean effective would be as follows : 

(i) Divide the mean effective for the H.P. cylinder by the 
ratio between the L.P. and H.P. piston areas. This reduces tEe 


H.P. mean effective to the L. P. piston. 

(2) Divide the mean effective for the I. P. cylinder by the 
ratio between the 1 L.P. and I. P. piston areas. This reduces the 
I.P. mean effective to the L.P. piston. 

(3) Add together the results found in (i) and (2) and the 
mean effective for the L. P. The sum will be the total mean ef- 
fective reduced to the L.P. piston. 


Given for a triple expansion engine the following: 

Diam. H.P. cylinder = 24" 

I.P. = 38" 

L.P. " = 60" 

Length of stroke = 42" 

Revolutions = 106 

From sets of indicator cards suppose the mean effective 
pressures found as follows : 
For the H.P. /* = 61.9 
" " I.P. p> = 30.2 
" " L.P. p* 13.8 

Find the total I. H. P. and the reduced mean effective pres- 

For the piston areas we have from a table of areas of 
circles : 

A l = 452.4 
A a =: 1134.1 
A 3 = 2827.4 

Then finding the I H. P. in each cylinder we have as 
follows : 

I. H. P. in H. P. cylinder = ........ 629.6 

I. H. P. in I. P. cylinder = ......... 770.2 

I. H. P. in L. P. cylinder = ......... 877.2 

Total I. H. P. ................ 2277.0 

Then according to rule (i) for the reduced mean effective : 

33,000 x 2277 g2 

1 7 X 106 X 2827.4 = 

According to rule (2) for the same we should have as 
follows : 

. 1134.1 

= 6l '9 >' + 30.* ! -~- + 13.8 

or/ = 9.9 + 12.12 + 13.8 = 35.82. 


The results are of course the same, since the two operations 
are simply two methods of computing the same quantity. If the 
I.H.P., revolutions, length of stroke and L.P. piston area are 
given, then the first method would be used. If the mean effec- 
tives in the various cylinders are given, together with the revo- 
lutions, length of stroke, and piston areas, then the second 
method may be used without necessarily finding the I. H. P. 
at all. 

Problems : 

(1) Given I. H. P. = 5,120. 
L. P. area = 5,612. 
Stroke 48". 
Revolutions = 112. 

Find the reduced mean effective pressure. Ans. 33.6. 

(2) From a pair of H. P. indicator cards the mean effective 
pressure is found to be 72.6 Ib. The diameters of the H. P. and 
L. P. are respectively 18 and 48 inches. Find the high pressure 
mean effective reduced to the L. P. piston. Ans. 10.2. 

(3) In the same engine as in (2) the mean effective pressure 
for the I. P. cylinder is found to be 33.2 Ib., and the L. P. diam. 
is 29 inches. Find the I. P. mean effective reduced to the L. P. 
piston. Ans. 12.1. 

(4) In the same engine as in (2) the L. P. mean effective 
pressure is 14.1 Ib. Find the entire reduced mean effective pres- 
sure. Ans. 36.4. 


The load on the crosshead guides comes from the load on 
the connecting rod and the obliquity of its line of action. The 
mechanics of this problem is considered in Part II., Sec. 12 [14] 
(15) and the maximum value of the load, which is found when the 
crank is at right angles to the center line, is readily computed 
in the manner there shown. It thus appears that the maximum 
load on the guide will bear the same relation to the load on the 
piston that the length of crank does to the connecting rod. This 
method of computing the load will be illustrated by an example. 

(i) At about mid stroke given the pressure on the top of 
the piston 180 Ib. per square inch and on the bottom 88 Ib. per 
square inch. The ratio of connecting rod to crank is 4.5 to i. 
The area of the piston is 404 square inches. Required the maxi- 
mum load on the guide. 


Net pressure on the piston =. 180 88 = 92 Ib. per square 

Net load on the piston = 92 X 404 = 37,i68 Ibs. 

Maximum load on guide 37,168 H- 4.5 = 8,260 Ibs. 

The safe load on guides is usually taken at from 50 to 70 Ibs. 
per square inch. In this case therefore taking 60 Ibs. as a safe 
load per square inch we should have : 

Area needed = 8,260 ~ 60 138 square inches. 


The net load on a slide valve is the difference between the 
steam loads on the two sides. On the back we have a load clue 
to the full steam pressure in the steam chest. On the inside we 
have a more variable load due partly to the pressure in the 
steam chest or cylinder, and partly to the exhaust pressure. For 
the low pressure cylinder exhausting into the condenser the 
exhaust pressure is small and is usually neglected. The area of 
the face subjected to pressure from the cylinder is also relatively 
small, and for the purposes we have now in view is usually 
omitted. The load on such a valve is therefore taken simply as 
the load on the back, the pressure per square inch, multiplied 
by the area in square inches. Denote the pressure by p and the 
area by A. Then the total load will be pA. The resistance to 
the motion of the valve which must be overcome by the valve 
rod will be the load pA multiplied by the coefficient of the fric- 
tion between the valve and its seat. Let f denote this coefficient, 
and F the force necessary to move the valve. Then we have : 


The values of f will depend on the condition of the surfaces 
and on the lubrication. With well fitted and lubricated surfaces 
its value should not exceed .01 to .02. With dry surfaces, es- 
pecially if they should begin to abrade, its value may rise to .10 
and more. 


Given a low pressure slide valve with dimensions 50 inches 
by 60 inches : average excess of pressure in valve chest over 
condenser, 26 Ibs. Coefficient of friction .02. Find load on valve 

Area = A 50 X 60 = 3,000 square inches. 

Load = pA = 26 X 3,000 = 78,000. 

Load on valve stem = fpA = .02 X 78,000 = 1,560 Ib. 


In designing a valve stem relative to such a load it must 
be given a large factor of safety in order to provide for starting 
the valve from rest or where partly stuck to the seat, and also 
for extra stresses due to the effects of inertia. 

For a flat slide valve on a high or intermediate cylinder, an 
estimate must be made of the load on each side and the differ- 
ence taken. Without serious error the net pressure may be 
taken as the difference between the average pressure in the 
valve chest and in the next following receiver. If then pi and 
/>* are these pressures, (pi p*) will be the difference, and 
(pi />*) A the average load on the valve. The remainder of the 
operation is, of course, the same as explained above for the 
low pressure valve. 


In Sec. 57 we have seen how to find the amount of heat 
required to formv one pound of steam of given temperature and 
pressure from a pound of feed-water of given temperature. To 
condense the steam and reduce it back to the condition of the 
feed-water will require the subtraction of the same amount of 
heat. Hence we may find the heat to be taken from each pound 
of steam in the condenser in exactly the same manner as in Sec. 
57. Now suppose the condensing water as it comes in to have 
a temperature of ft, and as it is discharged, a temperature of t*. 
Then the temperature of each pound will be raised (t* fi) de- 
grees. This means that it will absorb (t* h) units of heat. 
Then if we divide this into the number of heat units which must 
be taken from each pound of steam, it will give the number of 
pounds of condensing water which must be provided to con- 
dense one pound of steam. Then if we know the amount of 
steam to be condensed, the total amount of condensing water is 
readily found. 

This may be illustrated by the following example : 

Pressure of steam at exhaust = 3.5 pounds, absolute. 

Corresponding temperature = 148. 

Temperature of condensed water = 130. 

Temperature of condensing water at entrance or h == 62. 

Temperature of condensing water at discharge or t* 98. 

Then from the steam tables we find that 1,029 heat units 
per pound must be subtracted in order to condense the steam 
and reduce it to the condition of the water in the condenser. 


We have also t* U = 98 62 = 36 = number of heat units 
absorbed per pound of condensing water. Then 1,029 -f- 36 = 
28.6 = number of pounds of condensing water per pound of 

Let us suppose an engine of 2,000 I. H. P. requiring 16 Ibs. 
of steam per I. H. P. per hour to be condensed under these con- 
ditions. Then for the total weight of water W we have : 

W = 2,000 X 16 X 28.6 = 915,200 pounds per hour. 

And 915,200 -.- 60 = Ibs. per mt. = 15,253. 

Then 15,253 -f- 64 = cu. ft. sea water per mt. = 238. 

In all ordinary cases the number of heat units to be sub- 
tracted will not differ much from 1,000, and for a rough estimate 
this number is often taken without detailed computation from 
the steam tables. Then, varying with the season of the year 
and the locality, we may expect that each pound of condensing 
water will absorb from say 25 to 50 heat units, and hence that the 
condensing water required per pound of steam will vary from 
40 to 20. 


It is sometimes desired to find the net work done by a pump 
in handling a certain amount of water. This may be computed 
closely if we know the conditions under which the pump op- 
erates. It is shown in mechanics that work may be divided into 
a volume factor and a pressure per unit area factor, and this 
form of the expression for work is usually most convenient for 
use in such cases. 

Let us take first the case of a boiler feed pump feeding 
against a gauge pressure of 160 pounds and supplying 16 
pounds water per I. H. P. per hour for 2,100 I. H. P. Then the 
amount of water supplied will be 2,100 X 16 = 33,6oo pounds 
per hour. This equals 33,600 -f- 60 = 560 pounds per minute. 
Taking 62.5 pounds per cu. ft. this will occupy a volume of 560 
-:- 62.5 = 8.96 cu. ft. This volume of water is pushed into the 
boiler against a total pressure of 160 + 14.7 or say 175 pounds 
per square inch or 175 X 144 = 25,200 pounds per square foot. 
Hence we have: Work per minute = 25,200 X 8.96 = 225,792 
ft. Ibs. Reducing this to horse power we have : H.P. = 225,792 -f- 
33,000 = 6.84. This is the net work, and assuming that there is 
no leakage or loss of steam. Actually there will be such a loss, 
raising the amount of water which the pump must deliver by 
from 5 to 10 per cent or more. 


Now between the steam cylinder where the total work is 
developed and the net delivered work as above determined, there 
is a series of losses. These may be classified as follows : 

(1) Loss due to the friction of the water in the pipes and 
to the inertia or resistance of the valves. These items form an 
extra resistance which must be overcome in addition to the 
regular pressure in the boiler. 

(2) Loss due to the friction of the pump itself. This like- 
wise forms an extra resistance as in (i). 

(3) Loss due to the slip of the pump. The pump plunger 
and valves are rarely tight and a certain amount of "slippage" to 
the water is sure to occur. It results that the volume displaced 
by the pump plunger will be greater than the volume delivered 
to the feed pipe, and the work to be done in the water cylinder 
will be increased in about the same ratio. The slip is quite a 
variable feature, being quite small with good workmanship and 
careful attention, and large under contrary conditions. With 
the usual run of boiler feed pumps, however, it will rarely be less 
than 5 per cent., and with lack of care may readily rise to from 
10 to 20 per cent. 

The sum of losses (i) and (2) will be found usually between 
15 and 20 per cent., and hence the sum of the total losses may 
be expected to vary between perhaps 20 or 25 and 40 per cent. 

In the present case, for illustration, assume the loss by leak- 
age of steam at joints, etc., as. 6 per cent. Then the water ac- 
tually delivered to the boiler will require a net work of 6.84 -'- 
.94 = 7.28. Assume the total losses between steam cylinder and 
feed pipes to be 33 per cent. Then the I. H. P. in the steam end 
will be 7.28 -f- .67 = 10.87, an d we should therefore expect that 
under moderate to fair conditions such a pump would require 
from 10 to ii I. H. P. With the pump plunger and valves leak- 
ing badly, stiff working parts and generally poor conditions, the 
amount will, of course, rise far above these figures. The full 
capacity of the feed pump would also be, of course, consider- 
ably above these values. We are here simply concerned with 
an estimate of the power actually required and the net power 
delivered under a given set of conditions. 

Again consider the case of a centrifugal pump for the 
same engine handling we will say 30 pounds condensing water 
per pound of steam condensed. Then the total amount of water 
handled per hour will be 2,100 X 16 X 30 = 1,008,000 pounds. 


In this case the resistance to be overcome is due chiefly to 
forcing the water through the condenser tubes. In some cases 
also the discharge outlet is slightly above the surface of the 
water and this additional lift increases the work to be done. 

The most convenient method of computation in this case is 
to estimate the total head equivalent to the resistance occasioned 
by the condenser tubes, and the lift of the water above the sur- 
face. This will be the total height of water which would pro- 
duce the same pressure as must be overcome by the pump at 
the tips of the vanes. In usual cases we may assume the head 
corresponding to the resistance in the condenser tubes at from 
4 to 5 feet. If the water is discharged at or below the water 
level this will then be the total head against which the pump 
works. In case, however, the pump draws from the bilge as 
when used for freeing the ship of water, then the total head will 
be the total lift plus the head due to the condenser tubes, and its 
value may rise in such cases to 20 feet and more. 

In such cases the work done is the product of the weight of 
water handled as the force or resistance factor, multiplied by the 
head as the distance factor. 

In the present case, assuming a total head of 6 feet we have : 
Work per mt. = 16,800 X 6 = 100,800 ft. Ibs. 

or H. P. = 100,800 -f- 33,000 = 3.05. 

The efficiency of such pumps is usually found between .30 
and .50. That is, between 50 and 70 per cent of the power de- 
veloped in the steam engine operating the pumps is lost, chiefly 
in the slip of the pump. Hence under fair conditions we may 
assume that 'this 3.05 H. P. will be about 40 per cent, of 
the I. H. P. of the engine. Hence the latter will be greater than 
3.05 in the ratio of 100 to 40 or 2^2 times. Hence we have : 

I. H. P. required = about 2*/ 2 X 3.05 = about 7.5. 

These examples will serve to show the methods to be used 
in working such problems, and if the principles involved are kept 
clearly in view they may be similarly applied to the solution of 
many problems likely to present themselves in engineering work. 


It may be sometimes convenient to be able to compute ap- 
proximately the amount of steam which will escape into the 
atmosphere from a chamber under a given pressure through an 
aperture of given area. 


Let p be the pressure, supposed to be not less than 25 Ibs., 
absolute. Let A = area of aperture in square inches, and W = 
weight discharged in pounds per second. Then Napier's rule 
for the approximate value of W is as follows : 


Thus given A = .5 sq. in. and p = 140 we have : 

W = I4 X ' = i lb. per second. 
7 X 2 

Take the following problem : What weight of steam would 
be discharged per hour through a small hole or crack of area 
.005 sq. in. under a pressure of 200 pounds per sq. in.? 

Using the formula we have : 

TJ _ 200 x -005 x 3600 - , 

W = - - - 51.4 pounds per hour. 

We may thus realize the importance of small leaks. 


The determination of the weights of various parts of marine 
engines and boilers is often necessary as a part of an estimate 
of costs for repairs or for other purposes. Such determinations 
are usually made by numerical computation, and consist in find- 
ing first the volume of the piece in question, and then its weight 
by the use of factors such as those given in the Table on p. 30. 
The chief part of the computation is therefore mensuration, the 
principles of which are given in Part II., Sec. 9. We will here 
add some general suggestions regarding the application of these 
rules, with some additional methods which may be used in 
special cases. 

[i] Units to be Used. 

The dimensions will usually be taken either from a drawing 
or directly from the piece in question. It may be recommended 
as a general rule to reduce all dimensions to inches as the unit 
rather than feet or feet and inches, the latter requiring the use 
of duo-decimal notation and methods as explained in Part II., 
Sec. 4 [5]. Where the pieces are small and fractions of an inch 
are to be dealt with, it will usually be most convenient to re- 
duce them to decimal form rather than to express them as com- 
mon fractions. In brief the inch as the unit and numbers ex- 
pressed decimally are recommended as a general rule in such 


computations. The factors for reducing cubic inches to pounds 
are then used as given in the Table on p. 30, and the weight is 
readily found. 

[2] Approximations and Short Cuts. 

In computations of this character absolute accuracy does 
not exist. In fact with no physical measurement can absolute 
accuracy be attained. In practical life we wish simply an ap- 
proximation, a value sufficiently near for the purposes in view; 
a value so near that the error is not of commercial or financial 
importance. All engineering measurements and computations 
recognize this principle, and of the acquirements which may 
come to the engineer with experience, none is of greater value 
than that which enables him to know where to stop his compu- 
tations, how far to carry his measurements and approximations, 
what error will be of importance, and what insignificant. Thus 
if we are to measure the dimensions of a coal bunker in order 
to compute its volume, it is evidently absurd to note the figures 
to the fraction of an inch. We can be by no means sure that 
the length, for example, is uniform within any such limit, and 
the difference due to a variation of ^4 r Y^ inch either way will 
be insignificant for the purpose in view. On the other hand, if 
we are to measure a journal in order to make a new one which 
shall fit in the same bearing, the admissible error is only a few 
thousandths of an inch, and the utmost attainable accuracy will 
be in order. So likewise if we are finding the weight of a sheet 
of boiler plate . an error of l /% inch in the length or breadth 
will introduce no significant error in the final result, while such 
an error in the thickness would cause a most serious error in the 
result. The former would make a difference of perhaps one 
part in 1,000 or so, while the latter might cause an error of one 
part in 8 or 10. 

Another point which may also be remembered is that a large 
relative or percentage error is more permissible in some small 
part of the whole than in a large part. Thus in a boiler an error of 
10 per cent in the weight of the tubes may be of less importance 
than one of i per cent in the weight of the shell and heads, while 
an error of 50 per cent in the high pressure cylinder cover, for 
example, might make less difference than one of 10 per cent in 
the low pressure cover. Of course there should be no excuse for 
making any 50 or 10 per cent errors, but the principle may be 


borne in mind as a legitimate means of saving time when a 
roughly approximate value must be determined. 

The most common approximations are those which make 
the computation of volume simpler by substituting for the actual 
body some other of simpler form, and with such dimensions as 

Fig. 284. Approximate Area of Segment of Circle. 

Marine Enyintering 

Fig. 285. Approximate Area of Boiler Plate. 

to be of equal volume so far as judgment may be able to deter- 
mine. Such substitutions are often employed, but they must be 
used with judgment and care in order that the possible error in- 
troduced may not be larger than permissible. No general rules 


can be given for such approximations, but the most common 
consist in substituting a rectangle or triangle or sometimes a cir- 
cle, for a more irregular area ; or a cylinder or regular prism or 
plate for some irregular volume. 

Thus in Fig. 284, if we wish to find quickly an approximate 
value of the segment of the circle ABC, we may sketch in a tri- 
angle ADC, so taking the sides that the area left out shall be 
judged equal to that taken in, and hence the area of the triangle 
may be taken as an approximation to that of the segment. This 
area is then readily found by the usual rule for a triangle. 

Again, in computing the area of a front boiler tube sheet, as 
shown in Fig. 285, we may for a first approximation substitute 
by judgment for the actual contour a rectangle ABCD, and thus 
quickly obtain a value which may be sufficiently close for the 
purpose in hand. 

Again we may often add by judgment something to one of 
the dimensions of a piece in order to provide for additional or 
irregular parts which would not be included in the regular geo- 
metrical figures dealt with. Thus in finding quickly the approxi- 
mate weight of a cylinder casting, provision may be made for 
the flanges by adding by judgment an appropriate amoilnt to the 
length of the casting; or similarly for a piece of shafting with 
flanged couplings at the ends. 

It is often necessary to divide a more or less complicated piece 
into several parts, each of which may be of some relatively sim- 
ple form. In some cases the volume of one simple form may 
be subtracted from that of another, thus giving as the remainder 
the volume of a more or less irregular form. Thus, to find the 
volume of a pair of brasses with square backs and sides, we may 
find the volume from the outside dimensions as though the block 
were solid, and then the volume of the cylindrical hole, and take 
the one from the other. 

Many such little devices will suggest themselves in connec- 
tion with the details of the work ; but it will be unnecessary to 
here enter further into the subject. 

In connection with the rule of Pappus, Part II., Sec. 9 [30], 
we may note the following method of applying it to the deter- 
mination of the weight of such forms as a piston, cylinder head, 
etc. The operations are as follows : 

(1) The cross sectional drawing is supposed to be at hand. 

(2) A copy of the half cross section, as shown for a piston 



in Fig. 286, is prepared on thick, uniform paper, and then cut 
carefully out with a sharp-pointed penknife. 

(3) This is weighed on delicate scales, and also balanced on 
the knife edge, the line AB containing the center of gravity be- 
ing thus found. 

(4) A square of the paper containing any convenient num- 
ber, say 100 square inches of area, is also cut out and weighed. 
This divided by the area will give the weight of the paper per 
square inch. 

(5) The weight of the paper half section is divided by that 

Fig. 286. Volume of Piston by Rule of Pappus. 

of the square inch. The quotient will be the area of the paper 
half section in square inches. 

(6) This area is multiplied by the square of the scale ratio of 
the drawing. Thus if the drawing is to a scale I inch = I foot, 
it is in the ratio i : 12 with the original, and we multiply by 
12 X 12 or 144. If the scale is i^ inches = i foot it is i : 8, and 
we multiply by 64. If to a scale of 3 inches = i foot it is i : 4 
and we multiply by 16. The result thus found will be the area of 
the actual full-sized half section in square inches. 

(7) We then multiply the distance AG scaled off according 
to the scale of the drawing and expressed in inches, by 6.2832, 
and the product by the area as found in (6). The result will be 
the volume in cubic inches. 


Instead of the preceding, we may less accurately find the 
area by taking it in parts and using substituted simpler forms, as 
above explained. We may then by judgment assume the loca- 
tion of G and then proceed as above in (7). 

Thus, for example, suppose we find as follows : 
Scale of drawing i l / 2 inches = i foot. 
Weight of paper section 240 grains. 
Weight of paper per square inch 36 grains. 
Arm AG 1.7 inches on the paper or 13.6, as scaled from the 


Then area =. 240 -=- 36 = 6.67 inches. 
This multiplied by 64 gives 426.7 square inches as the area of the 

actual half section. 
Then volume 13.6 X 6.2832 X 426.7 = 36462 cubic inches. 




Sec. 78. MEASURE OF 

For measuring the speed of steamships the customary unit 
is the knot. While this term is often used as a distance, it is 
really a speed or velocity. As adopted by the United States 
Navy Department, it is a speed of 6080.27 ft. per hr. The Brit- 
ish Admiralty knot is a speed of 6080 ft. per hr. For all ordi- 
nary purposes the United States and British knots may be con- 
sidered the same. It is often necessary to reduce knots to feet 
per minute or vice versa. To this end we divide 6080 by 60 and 
find 101.33 ft. per minute as the equivalent of one knot. Hence 
the following rules : 

To reduce knots to feet per minute multiply by 101.33. 

To reduce feet per minute to knots, divide by 101.33. 

In the inland waters of the United States, and to some ex- 
tent on the coast for tugs, yachts, launches, etc., the mile per hour 
is used as the unit, instead of the knot. A statute mile consists 
of 5,280 feet. Hence one mile per hour equals 5,280 -r- 60, or 88 
feet per minute. Hence : 

To reduce miles per hour to feet per minute, multiply by 88. 

To reduce feet per minute to miles per hour, divide by 88. 


To reduce miles per hour to knots, divide the former by 


To reduce knots to miles per hour, multiply the former by 


To propel a ship through the water some kind of a propul- 
sive thrust must be obtained. This is the fundamental problem 
of propulsion. Thus when a boat is poled along a shallow creek 


the thrust is obtained as a reaction from the bed of the creek 
against the end of the pole and thence to the man who is pushing 
it, and thence to the boat. 

In the usual case, however, there is no bottom to be 
reached, and the only thing outside the ship which can be gotten 
hold of for the purpose of gaining a reaction is the air or the 
water. For all cases with which the engineer is concerned it 
comes to the latter, and so the problem is to get from the water 
a reaction or force directed forward by means of which the ship 
maybe pushed through the water. To understand how this is 
possible we must remember the property of inertia, one of the 
fundamental properties of matter. It is this property which en- 
ables all matter to resist any effort made to change its condition 
bf rest or relative motion, and to react back on the means by 
which such a change is effected. Thus a push of the hand may 
serve to set in motion a weight hanging by a rope, but while the 
condition of rest is being overcome the weight will react back on 
the hand with a force equal and opposite to that which the hand 
exerts upon the weight. Similarly, when a shot is fired from a 
gun, the inertia of the shot causes it to react back against the 
as and so to the gun, causing the well known recoil. 

From the fundamental principles of mechanics it follows 
that to obtain a thrust or reaction forward it is only necessary to 
produce in a certain mass of water an increase in velocity, such 
increase being directed from forward aft, or at least having a 
component in that direction. There will then be a reaction 
directed from aft forward, or having a component in that direc- 
tion, and such reaction exerted on the means used to produce the 
change of velocity may be utilized as a propulsive thrust. 

This is carried out in practice by either a screw propeller or 
paddle wheel, and remembering the principles above stated, it 
appears that the immediate purpose of the propeller or paddle 
wheel is simply to produce an increase in the velocity of the 
water directed from forward aft. In consequence of this the 
water will exert a forward reaction on the propeller or paddle 
wheel, and thus produce the thrust required to propel the ship 
through the water. It may be added that this increase of ve- 
locity from forward aft, referred to above, may be obtained either 
by taking hold of water at rest and giving it a motion stern ward, 
by taking hold of water already moving stermvard and giving it 
a still higher velocity in the same direction, or by taking hold of 



water moving forward and decreasing such forward motion, 
stopping it and leaving it at rest, or reversing it to a sternward 
motion. In all cases it is the change of velocity which is of im- 
portance. Thus a change from rest to 5 feet per second aft, or 
from 3 feet per second aft to 8 feet per second aft, or from 5 feet 
per second forward to rest, or from 2 feet per second forward to 
3 feet per second aft, will each give the same forward thrust. 

We shall not go further into the theory of propulsion, but in 
the next section will give certain definitions relating to screw 
propellers and the solution of a few simple problems. 

Sec. 80. SCREW 

[i] Definitions. 

A screw propeller as shown in the frontispiece and in Figs. 
287, 288, consists of a hub and a certain number of blades, 
usually two, three or four. The blades have on their rear or 
driving side an approximately helical surface that is, a surface 
similar to that which forms the faces of an ordinary screw thread. 
In this view a two-bladed screw propeller may be considered as 
a small part of a double-threaded bolt, the threads being cut very 
deep, and all portions being cut away down to the hub except 
the parts retained for blades. Similarly a three or four-bladed 
propeller may be considered as a small part of a triple or quad- 
ruple-threaded bolt similarly cut away except for the parts re- 
tained for blades. 

The hub or boss is the central portion to which the blades are 
attached, and through which they receive their motion of rota- 
tion in a transverse plane relative to the ship. 

A propeller is said to be right hand or left hand, according 
as it turns with or against the hands of a watch when looked at 
from aft and driving the ship ahead. 

The face or driving face of a blade is to the rear. It is that 
face which acts on the water and so receives the forward thrust. 

The back of a blade is therefore on the forward side. Care 
must be taken not to confuse these terms. 

The leading and following edges of a blade are respectively 
the forward and after edges. 

The diameter of a propeller is the diameter of the circle 
swept by the tips of the blades. 

The pitch of a propeller is the same as the pitch of the screw 
thread, of which it may be considered as forming a small part 



that is, it is the longitudinal distance between the successive turns 
of the helical surface. This definition will hold, however, only 
when the pitch is the same over the entire face of the blade. In 
many cases the pitch varies from one point to another, and we 
must therefore understand the term as relating, in such cases, 
to a small element of the driving face only. From this view the 
pitch may be defined as the longitudinal distance which the ship 
would be driven for one revolution were this element to work 

Fig. 288. Screw Propeller,, Detachable Blades. 

on a smooth, unyielding surface, as, for example, the cor- 
responding surface of a fixed nut. The pitch thus defined will 
depend on the location and inclination of the surface at the point 
or element considered, and its value may thus vary from one 
point to another over the entire face of the blade. The pitch is 
thus said to be uniform or variable, as its value remains the same 
or changes from point to point over the driving face. If it in- 
creases as we go from the hub to the tip of the blade it is said to 


increase radially. If it is greater on the following than on the 
leading edge, it is said to increase axially. The latter is usually 
implied by the simple term increasing or expanding pitch. 

The pitch ratio is the pitch divided by the diameter, or the 
ratio of pitch to diameter. 

The area, developed area or helicoidal area of a blade is the 
actual surface of the driving face. For the propeller as a whole 
it is the sum of the areas of all the blades. 

The projected area is likewise the area of the projection on 
a transverse plane, of one blade or of all the blades collectively. 

The disk area is the area of the circle swept by the tips of 
the blades. 

The pitch has been defined as the distance which the pro- 
peller in one revolution would drive the ship if it worked on a 
smooth, unyielding surface. Instead of working on such a sur- 
face, however, the propeller works on the water, a yielding me- 
dium, and in consequence the water recedes somewhat under the 
action of the propeller and the ship moves forward per revolu- 
tion a distance less than the pitch. The difference between the 
pitch and the distance the ship actually moves per revolution is 
called the slip, or, more concisely, the slip per revolution. The 
ratio of this slip to the pitch is called the slip ratio, or simply the 
slip stated in per cent as a slip of 20 per cent, 30 per cent, etc. 

Let p denote the pitch of the propeller in feet. 
N " " revolutions per minute. 
u " ' velocity of the ship in knots. 
^ " " slip ratio. 

Then 101.3 -r- AT is the distance traveled by the ship per 
revolution, and p (101.3 w N) is the slip per revolution. 
Hence for the slip ratio we have : 

101.3 " 

Multiplying both terms of the fraction by ^V we have: 

AY- ""-a K ,. 


The term pN is the distance the ship would go per minute if 
there were no slip, while 101.3 is the distance which is actually 
made good. The difference or pN 101.3 u may therefore be 
called the slip per minute, and the quotient of this by /\Y is the 


slip ratio. This latter equation for ^ is the one by means of 
which its value is usually computed. 

It may be well to note at this point that while slip implies a 
certain loss of effectiveness in the propeller, it is a loss which is 
necessary in the very nature of the case. We have already seen 
that to obtain a propulsive thrust we must give to a certain body 
of water an increased velocity sternward. This means that the 
water must yield under the action of the propeller, and it is this 
yielding or falling sternward which thus gives rise at the same 
time to both the slip and the propulsive thrust. We cannot 
therefore have the thrust without the slip: we must accept the 
latter to obtain the former. 

We must now introduce a further consideration. We have 
defined slip as the difference between the pitch on the driving 
face and the advance per revolution. The latter admits of being 
defined in two ways according as we take for our point of refer- 
ence a point in the outlying still water, or a point in the water 
about the stern of the ship and in which the propeller works. So 
far as we are concerned with the movement of the ship through 
the water as a whole, the former is the natural point of reference. 
For various considerations connected with the operation of the 
propeller itself, however, the latter is the more important. Let 
us note briefly the condition of the water close about the stern of 
the ship and in which the propeller works. 

The ship in moving through the water will throw into for- 
ward motion a skin of water extending from the surface of the 
ship for several inches outward. Very near the surface of the 
ship this will move with nearly the velocity of the ship, while as 
the distance from the surface is increased the velocity will rap- 
idly decrease and soon become insensible. The water thus given 
forward motion by the skin of the ship will finally be found at 
the stern, where, still further influenced by wave and stream line 
motion, it forms the so-called "wake." The forward velocity in 
the wake at different points in a transverse plane at the stern is 
quite irregular, rising as high as 50 to 75 per cent of that of the 
ship at points near the surface and near the stern post, and de- 
creasing irregularly and gradually to nothing at the outlying still 
water. For single screw ships the average value in that part of 
the wake directly influenced by the screw is usually from 10 to 
20 per cent of the speed of the ship. For twin screws located 
somewhat aside from the strongest part of the wake the values 


are usually found between 6 and 12 per cent of the speed of the 

Now with reference to the propeller, it is evident that so far 
as it is concerned individually and as an appliance for developing 
thrust, it should be judged relative to the water immediately 
about it and in which it works rather than relative to an outly- 
ing body of undisturbed water upon which it has no direct influ- 
ence. The slip which is given by taking the speed relative to the 
wake is therefore called the true slip, while that given by taking 
the speed relative to the outlying still water (the speed as usually 
considered) is called the apparent slip. 

To show the relation between the true and apparent slips let 
v denote the forward velocity of the wake and u that of the ship 
as before, both measured in knots and relative to the outlying 
still water. Then (u v) is the speed of the advance of the pro- 
peller through or relative to the wake. Also using the same no- 
tation as above, 

p N 101.3 (u v) is the true slip per minute, while, as before, 
'p N 101.3 u is the apparent slip per minute. Denoting the lat- 
ter by 5"i, and the former 5"*, we have : 

Si = p N 101.3 u. 

S* = p N 101.3 (u v) = 5"i + 101.3 v. 

It thus appears that the difference between the two slips in 
feet per minute is simply the wake velocity, as we should expect. 

To reduce to slip ratio we use pN as the divisor in each case, 
and denoting the resulting ratios by si and s* we have : 

pN 101.3 u . 

s t = - - as before, and: 


fNi6i*3(u v) 101.3 v 

~pN~ = S * + pjy 

Where the term slip is used without special definition, the 
apparent slip is usually intended. 

In the preceding discussion of slip we have used the term 
pitch as though it were of constant value over the entire surface 
of the blade. If such is not the case, then the term must be un- 
derstood as referring to a mean or average value. Such an aver- 
age value of the pitch of a propeller might be defined in a variety 
of ways, but engineers are not as yet agreed upon the method 
most suitable. This point, however, is one which cannot be fur- 
ther developed in the present work. 



(1) To find the apparent slip. 

From the preceding value of the apparent slip ratio s we 
may derive the following : 

Rule : (i) Multiply the pitch by the revolutions per 

(2) Multiply the speed in knots by 101.3. 

(3) Subtract the result in (2) from that in (i) and divide the 
difference by the result in (i). The quotient will be the slip ratio. 

Example : Given speed of ship, 1 1 knots ; pitch, 20 feet ; 
revolutions, 72. Find the apparent slip. 
Operation : 20 X 72 = 1440. 

ii X 101.3=1114.3. 

1440 1114-3 -i- 1440 = 3 2 57 -=- J 440 = 22.6 per 
cent, ans. 

(2) To find the speed, having given the other items. 

From the preceding equation we derive the following value 
for u: 

u . >" (' - '> (,) 


Whence the following : 
Rule: (i) Multiply the pitch by the revolutions. 

(2) Subtract the slip per cent from i .00. 

(3) Multiply the result in (i) by that in (2). 

(4) Divide the result in (3) by 101.3. The quotient will 
be the speed in knots. 

Example: Given pitch, 18 feet; revolutions, no; apparent 
slip, 18 per cent. Find the speed. 
Operation : 18 X no = 1980. 

i.oo .18 = .82. 

.82 X 1980 = 1624. 

1624 -T- 101.3 = 16.03 knots, ans. 

(3) To find the revolutions, having given the other items. 
From the preceding equation we derive the following value 

for N: 

,, 101.^ U 

* ' 

The use of this formula will be illustrated by the following 
example : 

Given speed, 22 knots; pitch, 26 feet; apparent slip, 16 per 
cent. Find revolutions. 


Operation : 101.3 X 22 = 2228.6 

(1.00 .16) X 28 = 23.52 

2228.6 -f- 23.52 = 94.8 revolutions per minute, ans. 
(4) To find the pitch, having given the other items. 
From the preceding equation we derive the following value 
for p. 

101.3 if 

P - -N(is) < 4 > 

The use of this formula will be illustrated by the following 
example : 

Given speed, 28 knots; apparent slip, 21 per cent; revolu- 
tions, 380. Find the corresponding pitch. 
Operation : 101.3 X 28 = 2836.4 

380 X (i.oo .21) = 300.2 
2836.4 -4- 300.2 = 9.45 feet, ans. 

[3] Varieties of Propellers. 

Screw propellers are found in the greatest variety according 
to the number, shape, style and arrangement of blades. In mod- 
ern practice the number of blades is usually either three or four, 
the former being perhaps more commonly met with in twin 
screws and the latter in single screws. 

The shape of the blades may be oval or elliptical, as in Figs. 
287, 288, or broadening somewhat toward the tip with rounded 
corners, or of any intermediate or similar form which may be 
desired. The oval or generally rounded form of blade is most 
commonly met with in modern practice. The blades may also 
be bent or curved in various ways. Thus in propellers for small 
boats the blades are often bent back, as in Fig. 287, so as to 
throw them somewhat farther from the stern post. They are 
also sometimes curved in the plane of rotation so that the en- 
tering edge is well rounded as it enters the water in going- 
ahead. Combined with these there may be various modifications 
of pitch, as above referred to. The normal or standard pro- 
peller in modern practice may, however, be considered as one 
having plain blades of uniform pitch, of oval or elliptical form, 
and standing at right angles to the axis, as in Fig. 288. Most 
of the variations from this type are based on fancy rather than 
on definite engineering reasons. So far as is known at present, 
the simple normal type, as specified above, is the equal of any 
of those of variable pitch or of special form or shape of blade. 


and while it may be that some special combination of pitch, 
shape and form of blade may give a higher efficiency than can be 
obtained from the normal type, yet up to the present time such 
results have not been proven. 

Small propellers are usually made complete or in one cast- 
ing, as in Fig. 287. Large propellers are made either in one 
casting or with separate or sectional blades, as in Fig. 288. In 
the latter case the root of the blade carries a circular flange fit- 
ting into a corresponding recess in the hub. This serves to se- 
cure the blade to the hub by stud-bolts passing through the 
flange and fitted with nuts countersunk below its outer face. 
The general details of this arrangement are shown in the figure. 
The holes in the flanges are usually made slightly oblong, thus 
providing for a slight change in the pitch by turning the flange 
back and forth, and thus changing the average obliquity of the 
blade to the axis of the propeller. Once adjusted as desired the 
holes are filled by packing pieces so that no further change can 
result from the accidental slipping of the flange under the nuts. 
In this connection it may be noted that the change of pitch re- 
sulting from such a twisting of the blade is not the same for all 
parts of the blade, but varies from root to tip. It follows, if the 
blade is made of uniform pitch, that it will remain so only sa 
long as it is set at the corresponding angle, and that if it is 
twisted to and fro the pitch will be increased and decreased, but 
not uniformly ; so that in all positions but this one the pitch will 
no longer be uniform, but variable. For a moderate angle of 
twist, however, the change from uniformity is but slight, and 
changes of average pitch up to perhaps 10 to 15 per cent may be 
made without serious departure from the average. 

The chief advantages of the separate blades lie in the possi- 
bility of varying or adjusting the pitch as just described, and in 
the readiness with which repairs may be executed. A separate 
blade broken or defective may be .readily removed and replaced 
with a new one, this operation in small vessels being sometimes 
accomplished without placing the vessel in dry dock. One or 
two blades may also be carried as spare parts or shipped by rail, 
or otherwise, much more readily than an entire propeller. 

The attachment of the propeller to the shaft is shown by the 
figures. The taper is usually about I inch in the diameter per 
foot. The propeller is prevented from turning on the shaft by 
one or more keys fitted, as shown. The after end of the shaft 


is fitted with a nut which serves to hold the propeller against 
any tendency to slip off when backing, and this is often covered 
with a conical tail piece, as shown, in order to reduce the eddy 
formation just aft of the boss. It may be noted that for a right- 
hand propeller the nut is usually left hand, and vice versa, it be- 
ing considered that this arrangement reduces the liability of 
loosening or backing off. 

If water is allowed to come into contact with the taper of 
the shaft on which the propeller boss is secured it may give rise 
to considerable corrosion, and this may set the boss so firmly on 
the shaft that great difficulty will be experienced in its removal. 
In order to prevent the contact of water with the taper, various 
means may be employed. In one method the brass liner on the 
shaft is carried along nearly to the forward end of the taper, and 
a rubber ring is placed just forward of the boss or in a counter- 
bore. As the boss is forced on, the rubber ring is compressed 
between the boss and the liner, and thus a water-tight joint is 
made. See Fig. 288. In other cases the liner is carried into a 
counterbore in the boss and a red lead joint is made between 
the two. 

For the examination or repair of the stern bearing it may 
become necessary to remove the propeller and withdraw the tail 
shaft forward into the ship. This is of course an operation re- 
quiring the docking of the ship. For the removal of the pro- 
peller the first attempt may be made with steel wedges between 
the forward face of the boss and the stern post, or after end of 
the stern bearing. The space between the two is made up with 
the metal blocking necessary, and the wedges are then inserted 
and driven one from either side. In this way a tremendous 
strain can be exerted and the boss will be started unless seriously 
corroded or jammed unduly tight. In the latter cases recourse 
must be had to hydraulic jacks, or to a heavy ram or to heating 
the boss in order to expand it in size and break the connection 
with the shaft. Once the boss is started the weight of the pro- 
peller may be taken by chain hoists suspended from the counter, 
and the shaft may then be drawn forward into the tunnel. 

In the case of twin screw ships with the common form of 
strut, the same general means may be employed, except that it 
should be remembered that the strain set up is carried on the 
strut, and the means taken should not go so far as to endanger 
its rupture or the undue straining of its fastenings. 


[3] Materials. 

Cast iron, cast steel, brass, gun metal and the various bronzes 
are the materials used for screw propellers. Cast iron is the 
cheapest, but is relatively weak and brittle, and the blades must 
necessarily be thicker and less efficient than if made of steel or 
bronze. Cast steel is stronger than cast iron, and the sections 
may be accordingly decreased with a resultant gain in efficiency. 
The surface of cast steel is naturally not as smooth as with cast 
Iron, but with improved methods of production the difference is 
not important. Brass and the various bronzes have naturally 
a smoother surface, and seem furthermore to have a lower co- 
efficient of skin resistance. This, added to their strength and 
good casting qualities, makes possible a smooth and relatively 
thin blade with sharp edges, all of which are features favorable 
to good efficiency. With the best bronzes the ultimate strength 
may vary from 50,000 to 60,000 pounds per square inch of sec- 
tion. With cast steel the ultimate strength will reach still higher, 
or, say, to 65,000 pounds per square inch. With gun metal an 
ultimate strength of 25,000 to 35,000 pounds may be expected, 
while with common brass and cast iron not more than 20,000 to 
25,000 pounds can be depended on. 

Of the various materials available, manganese bronze may 
perhaps be considered as possessing the best combination of 
desirable qualities, such as strength and stiffness, good casting 
qualities, resistance to corrosion, etc. Care is needed in the 
manipulation of the various bronzes in melting, pouring and 
cooling, in order to insure uniformity and the full realization of 
the valuable properties of the alloy. The greater cost of such 
bronzes restricts their use, however, to warships, yachts and 
launches, ocean liners and other cases where the importance of 
a saving in propulsive efficiency may be considered worth the in- 
creased cost of the propeller. 

The durability of propeller blades is usually in the order: 
bronze, cast iron, cast steel. The two latter usually suffer by 
general corrosion and local pitting, the average life being usually 
from five to ten years. The life of bronze blades is practically 

[4] Measurement of Pitch. 

To determine the pitch of a given propeller three measure- 
ments are necessary. See Fig. 289. These are : 

(i) The radius OA at which the pitch is desired. 



(2) The angle or part of the complete circumference cor- 
responding to the distance on the blade between A and B, the 
two points between which the pitch is to be found. 

(3) The advance BC parallel to the line of the shaft, corre- 
sponding to this part of a complete revolution. 

In the figure, A and B are points on the face of the blade, 
and are at a constant distance OA from the shaft center line OO. 
AC is an arc of a circle which lies in a plane through A and per- 
pendicular to the shaft. The angle AOC is therefore the one re- 
ferred to in (2), and the distance BC is the corresponding ad- 
vance. Then BC is the same fraction of the entire pitch that 
AOC is of a complete circle, or the same fraction that the length 

Fig. 289. Measurement of Pitch. 

AC is of a complete circumference with OA as radius. This 
complete circumference will be 6.2832 X OA. Hence the pro- 
portion : 

AC : 6.2832 X OA : : BC : pitch. 

6.2832 x OA x BC 

or pitch 


It is not, however, as easy to measure AC as AB, so that 
we may put for AC its equal j/ ^-g *_ p^.and we then have 

pitch == 1^3*_x.OAjLgC 

AB 2 - h'C 2 
A brief outline of the operations is as follows : 


(1) Select the points A and B at and between which the 
pitch is desired, making sure that they are at equal distances 
from the shaft center line. This can be done by squaring down 
from a straight edge or other reference line PQ, PR, placed 
across the hub and at right angles with the shaft. Then measure 
the length AB. 

(2) The propeller being leveled up, measure the distance 
BC from a level through A vertically down to B. Or if the pro- 
peller cannot be leveled, measure from B in a direction parallel 
to the shaft out to a line through A in a plane at right angles to 
the shaft. Or measure from Q down to A and from R down to 
B and take their difference BC. 

(3) Multiply the distance OA or its equal PQ by 6.2832, and 
by the length BC all in the same units of measure. 

(4) Square the lengths AB and BC, subtract the square of 
the latter from that of the former and extract the square root 
of the difference. 

(5) Divide the result found in (3) by that found in (4) and 
the quotient will be the pitch desired. 

Thus, for example, suppose 

AB = 20 inches, BC = 13 inches and OA = 48 inches. 

Then 6.2832 X 48 X 13 39 2 Q-7 

AlSO 1/400 169 - 1/237 =: 15.2. 

Then 3920.7 -f- 15.2 = 238 inches = 21 ft. 6 in. = pitch. 

If the pitch is variable instead of uniform, the operation is 
precisely the same, but the result found must be considered 
merely as the mean or average value of the pitch between the 
points A and B. For other parts of the blade a similar process 
will give the pitch at those points. 

When the propeller is in place on the ship it is sometimes 
more convenient to carry out the principles involved in this 
method of measuring pitch somewhat differently, as follows : Let 
the propeller be turned so as to bring one of the blades horizon- 
tal. Then select the place at which the pitch is desired, and hang 
over the blade at this point a cord with two weights, as shown in 
Fig. 290. Care must be taken that the two points A and B at 
which the cord touches the edges of the blade are at the same 
distance from the center. It is then readily seen that the points 
A and B of Fig. 290 correspond to the similar points of Fig. 289, 
except that in Fig. 290 they are of necessity taken on the ex- 
treme edges of the blade. We then level up a bar PQ and meas- 



ure the distances, AB and BC, as noted above, using them in the 
same way for finding the pitch. Or we may measure AC directly 
and use this with BC in the proportion above. 

As a rough and ready rule it may be remembered that the 
pitch of a propeller will equal the length of a circumference at 
the place on the blade where the slope of the face is 45, or where 
it is equally inclined to the shaft and to the transverse direction. 
Starting near the shaft, the inclination to the longitudinal is 
small, but increases toward the tip, passing at some point 
through the value 45. At this point let the radius be r. Then 

Marine Enyinefri*g 

Fig. 290. Measurement of Pitch. 

pitch = 2 TIT = 6.2832 r. In this way an approximate idea may 
often be quickly obtained of the pitch of a wheel by estimate 
without special measurement, except for the radius or diameter 
at which the blade has the slope of 45. 

The details of the above methods for finding pitch may vary 
considerably, but the description given will serve to show the 
principles involved, and with reasonable mechanical skill no 
trouble will be found in carrying out the measurements required. 

Sec. 81. PADDLE 

In addition to the screw propeller the paddle wheel is the 
other appliance used for ship propulsion. In Fig. 291 is shown 
in skeleton a common radial paddle wheel. In this type of 
wheel the paddles or floats are rigidly fixed to the arms, the lat- 



ter being connected at their inner ends to a hub, which is carried 
on the shaft. In this manner the motion of the shaft is transmit- 
ted to the floats, and these, acting on the water, drive it stern- 
ward and thus receive the forward thrust which is required for 
the propulsion of the vessel. 

In Fig. 292 is shown a feathering paddle wheel. In this ar- 
rangement the floats are hung on axes and are swung in such 
way that they enter and leave the water nearly in an edgewise 
direction. In this way there is less disturbance of the water and 
a smoother action of the wheel is obtained. Such arrangement 
is especially suitable for ships operating under widely varying 
conditions of draft, for the floats of a deeply immersed radial 

Fig. 291. Radial Paddle Wheel, Skeleton Diagram. 

wheel enter and leave the water at a great obliquity and there 
would be considerable loss by oblique action. 

There are two chief methods by which the proper motion 
may be given to feathering floats, depending on whether the pad- 
dle shaft has an outer or spring bearing on the outside of the 
paddle box or is overhung ; that is, provided simply with a bear- 
ing on the rail, the paddle wheel itself being then mounted on 
the overhung end of the shaft. In the former case the arrange- 
ment will be understood from the skeleton drawing of Fig. 293. 
The stationary excentric A has its center forward of the wheel 
center, as shown. To the excentric strap is attached a drive link 
HB, connected by pin joint to an arm BC, carrying a float DE. 





The other floats, mounted in a similar manner, are connected by 
pin joint links to the excentric strap, as shown. As the wheel 
turns the drive link HB carries the strap around the excentric 
sheave, and with it the series of connected links. This gives a see- 
saw motion to the ends of the arms BC and thus swings the floats 
in the manner desired. 

When the paddle shaft has no outer bearing, as in the ar- 
rangement shown in Fig. 292, the disc carrying the links may be 
mounted on a supporting pin carried on the outer side of the 
guard. It may then be given motion through a drive link and 
connections, as shown, giving a similar see-saw motion to the 
floats, as in the former case. 

In modern practice the arms of paddle wheels are made of 

Fig. .293. Paddle Wheel, Skeleton of Arrangement for Feathering Floats. 

steel, the hubs of cast iron or cast steel, and the floats of wood 
or boiler plate ; in the latter case often curved in cross section. 

In estimating the pitch of the paddle wheel or what corre- 
sponds to pitch in the screw propeller, we must consider it as the 
circumference of the circle traveled by the floats. Since, how- 
ever, a float as a whole is made up of a series of strips or ele- 
ments at varying distances from the center, each such element 
will have its own circumference and therefore its own pitch, arid 
will try to drive the ship at a speed corresponding to such pitch. 
The paddle wheel as a whole has therefore a varying pitch, in- 
creasing from the outer to the inner edge of the float. The re- 


sultant mean pitch is considered as the circumference traveled by 
a point called the center of effort. The proper basis for the deter- 
mination of this point, and hence of the true mean pitch of a pad- 
dle wheel is, however, not definitely known, and can only be de- 
termined by the aid of extended experimental investigation. In 
the absence of such definite basis it is sufficient for all practical 
purposes to take it at the center of the float radially, though its 
true location would lie somewhat outside this point. Counting 
the circumference through this point as the pitch, the actual dis- 
tance traveled by the boat per revolution is less by the amount of 
the slip, which is usually found from 15 to 25 or 30 per cent. 

The circle whose circumference is equal to the distance trav- 
eled per revolution is sometimes known as the rolling circle. It 
is so called from the fact that the speed of the boat is the same as 
though it were carried on wheels of this diameter, "which rolled on 
a supporting surface as wagon wheels along a smooth, level road. 
The solutions of problems relating to the revolutions, diam- 
eter and slip of paddle wheels are found in the same general man- 
ner as for the screw propeller, and the re-suiting equations are 
similar to those found in Sec. 80 [i], with the substitution for p 
of D, as defined below, and 88 for 101.3. 

Let D = diameter of rolling or pitch circle. i 

N = revolutions per minute. 

u = speed in miles per hour. 

^ i= slip ratio. 

Then, as with the screw propeller, we have : 
DN - 88 u 

88 u 
D - 

These may be illustrated by the following examples : 

(i) Given diameter of rolling circle, 36 feet; revolutions, 45 
per minute ; speed, 14 miles per hour. Find the slip ratio. 
Operation : 

D N = 36 X 45 = 1620 

88 u 88 X H = 1232 

1620 1232 -i- 1620 = 388 -r- 1620 = 23.3 per cent. 


(2) Given diameter of rolling circle, 30 feet ; revolutions, 50 
per minute, what speed can be made,, allowing a slip of 30 per 

Operation : 

D N = 30 X 50 = J 5oo 

i s = i .30 = .70 

.70 X 1500= 1050 

1050 -r- 88 = 11.9 miles per hour = 10.35 knots. 

(3) Given a speed of 18 miles per hour, a slip of 24 per cent, 
and a wheel whose rolling circle has a diameter of 40 feet. Re- 
quired the number of revolutions per minute. 

Operation : 

88 X = 88 X 18 = 1604 

i ^ = i .24 = .76 

D (i s) = 40 X 76 = 304 

1604 -T- 30.4 = 52.8 revolutions per minute. 

(4) Given a speed of 14 miles per hour ; revolutions, 40 per 
minute ; slip, 28 per cent. Find the corresponding diameter of 
rolling circle. 

Operation : 

88 X = 88 X 14 = 1232 
i s= i .28 = .72 
N (i s) 40 X -72 = 28.8 
1232 28.8 = 42.8 feet. 


The subject of the pow r ering of ships is one which can be here 
only referred to in a brief and elementary way. The usual prob- 
lems are to find the power required to drive a given ship at a pro- 
posed speed, or the probable speed for a given ship with a given 
power. Such problems require a knowledge of the relation be- 
tween power, speed and the ship. In the present state of our in- 
formation on this subject, such relation cannot be accurately ex- 
pressed by any ordinary formula or equation. Several approxi- 
mate formulae have, however, been employed for the solution of 
such problems, and among them none has perhaps been of wider 
general usefulness than the so-called Admiralty coefficient 

Let/f= I.H.P. 

D = displacement in tons. 

v = speed in knots. 

K = a coefficient. 


Then according to this formula we have : 

H - * 

and solving for speed : 
3 /~HK~ 

and solving for the coefficient : 

K = j^_ 

The whole point in the use of the formula is to properly se- 
lect the values of the coefficient K in accordance with the special 
features of the case, including the form and size of the ship, pro- 
posed speed, probable efficiency of propulsion, etc. The safest 
plan is to find values of K from the trial data of actual ships of 
about the same size, character of form and speed, as the pro- 
posed case, and to be guided by such values in the selection of 
the coefficient for the proposed case. There are other special 
methods for obtaining from the trial data of ships of similar 
form, by the so-called law of comparison, the suitable values for 
a proposed case, even when the sizes and speeds differ consider- 
ably from those of the proposed case. Into the details of these 
points, however, we cannot here enter. Some general sugges- 
tions regarding the value of K with a few illustrative examples 
must suffice. 

We find then by experience that in general the value of K 
is greater (and hence the I.H.P. relatively less) as the ship is 
larger, but more especially as she is longer, also as she is nar- 
rower in proportion of length to beam, and as she is finer in 
form, especially in the water lines. 

In the reverse cases the values of K will be smaller, and the 
I.H.P. relatively larger. The values of K are also smaller and 
the I.H.P. relatively larger as the speed is higher in proportion 
to the length, or, more exactly, as the speed is higher in propor- 
tion to the square root of the length. For small launches and 
such craft driven at speeds in miles or knots greater than ;/Z i n 
feet, the values of K will be quite small, ranging perhaps from 
100 to 150. At lower speeds equal to or less than t/Z" the values 
will rise to perhaps 200 and more with fine form and^small pro- 
portion of beam to length. For yachts and craft of similar form, 
moderately fine and at fairly high speeds, values of 200 above 


and below will be found. For torpedo boats, with their narrow 
proportions and fine form, their excessive speeds carry